This document contains the literature review undertaken for DTFH61-11-F-00031, “Pedestrian and Bicycle Data Collection”. This document contains seven sections: Summary, Introduction, Equipment, Data Collection Programs, Data Processing and Storage, Data Use and Reporting. Each section contains a summary of the materials found, the potential gaps relating to conducting a bicycle and pedestrian counting program and an annotated bibliography organized by sub-topics.

The sources used included the Pedestrian and Bicycle Information Center (PBIC) online libraries, TRIS, ITRD (the database of the OECD's Joint Transportation Research Center), the ITE Journal, the APTA Journal and the TRR on-line database. The literature search focused on papers written in English. The review emphasized on research from the past decade to ensure that emerging technologies and methods are properly identified. Included in the search was literature discussing how various counting programs are conducted by local governments, MPOs, state DOTs, national, and international transportation agencies.

Information on bicycle and pedestrian data collection equipment and data collection programs is significant. Research papers and case studies describe a variety of manual and automatic counting methods. Additionally, new technologies such as GPS devices and smartphones are making bicycle and pedestrian data collection more dynamic.

The literature review also revealed that it is common to mix data and analysis methods to analyze data or extrapolate trends for planning purposes. From the review of data collection programs, it is common to combine manual count data, automatic count data and survey data to analyze bicycle and pedestrian activity.

Areas where information is lacking and in need of expansion include data processing and storage, data use, and reporting. Related to data processing and storage, the main information gaps are related to record formats and file structures. While it is clear from the literature that standards have been established through several programs, such as the National Bicycle and Pedestrian Documentation Project, information on the specifics of these topics was not found in the literature. The same information issues are related to reporting as well. Some agencies have developed standards for reporting. Other agencies either lack standards or have standards that are very different from those generally used.

While data use is documented in the literature, it is still clear that uniform standards for its processing and application are needed. Typically, data is used to analyze a local facility, is project specific or applies only to a specific geography. More information is needed on how bicycle and pedestrian data collection can be used to address safety, funding or legislative mandates, facility design and quality of service.

INTRODUCTION

Nonmotorized monitoring has become highly relevant to numerous agencies' planning efforts. However, there is no standardized technology for automated counts. Most bicycle and pedestrian monitoring programs use periodic manual counts. Continuous monitoring programs are becoming more commonplace with infrared counters as the most popular technology. Video and laser counting technology is promising. Pneumatic tubes, while not practical for pedestrian traffic, have been used effectively on cycle tracks. Each of these methods has its best applications given specific operational, geometric, and weather related factors. Currently limited information is available to agencies for standardization of non-motorized counting programs. Widely available documents on non-motorized facilities include the following:

The AASHTO Guide for the Planning, Design, and Operation of Pedestrian Facilities
The AASHTO Guide for the Development of Bicycle Facilities

Neither document contains more than a handful of sentences addressing count collection, utilization or storage.

This project is intended to provide a summary of the state of the practice for counting programs which collect data on non-motorized vehicles and pedestrians. The collection methods, equipment, factoring, storage, and reporting are all of interest. This material is being used to develop a set of webinars and provide input to the update of the Traffic Monitoring Guide (TMG). The webinars are intended to be a two-way exchange of information. A summary of this review will be presented to individuals involved in non-motorized data collection to increase awareness of the state of the practice. As a part of the webinars, unpublished information on current practice and activities will be sought. The summary of both activities will be available to update the TMG, including potential record formats which the community could adopt to simplify data exchange between organizations.

The review began with the identification and prioritization of sources including the Pedestrian and Bicycle Information Center (PBIC) online libraries, TRIS, ITRD (the database of the OECD's Joint Transportation Research Center), the ITE Journal, the APTA Journal and the TRR database. The literature search focused on papers written in English but looked at both U.S. and foreign practices. The review emphasized, but was not limited to, research from the past decade to ensure that emerging technologies and methods were properly identified. A second phase of the literature review looked at implementing jurisdictions' (local governments, MPOs, state DOTs, national and international transportation agencies) for application of the techniques. Listings were generated for both U.S. and international literature and then trimmed to those most likely to contain implementation rather than academic outcomes.

The following chapters are generally organized with an overview of the content, materials that only reference pedestrians, materials that only reference bicycles and then materials that either reference both or for which the difference in modes is not applicable. Not all chapters contain all subsections. Within each subsection the relevant sources are cited. Following the citation is a table indicating the relevant project topics for the material and a short summary of the material.

A major and not unexpected consistency found throughout the literature is that there is a lack of generally accepted standards for bicycle and pedestrian data collection. Several programs and case studies, such as the National Bicycle and Pedestrian Counting Project, have served as de facto standards. Substantive and vetted standards for bicycle and pedestrian data collection are still needed to provide consistent, multi-use/multi-user data of known quality and adequate quantity.

There is a demand for bicycle and pedestrian count data. The data is needed to inform policy, to support modeling and to assist with project design and implementation. However, a lack of resources is a barrier to meeting bicycle and pedestrian data collection needs.

In terms of equipment and methods, manual counts are still the predominant bicycle and pedestrian data collection method. Much of this can be attributed to the low capital costs required for manual counting and the fact that additional data, such as user characteristics, can be collected in addition to volume data. That being said, programs are increasingly using a blend of automatic and manual counting methods to collect data, with the method varying depending on the site context and information needed.

Beyond the common themes found in the literature, there are also several gaps in information and research that need to be filled. Notably, more comprehensive testing of automatic counting equipment is needed. The technology available today is analyzed sporadically and typically as part of a case study. A test method or other means to validate counting technology that practitioners can use to select the tools would be useful.

Another major gap in information is documentation of any widely accepted methods for data collection, storage and analysis. Data collection at all levels of government varies considerably from agency to agency. Some agencies have robust programs while others do not have a bicycle and pedestrian data collection program at all. This variance in the quality and quantity of data makes it difficult to transfer or compare bicycle and pedestrian data by geography, facility type, user, etc.

EQUIPMENT

Equipment may be divided into two categories: point-based and individual based. Point-based equipment covers sensors installed in, along or above the path for which volume counts are to be collected. Individual based equipment moves with the user to capture a series of locations in time and space from which trips may be inferred.

From the literature review, manual counting appears to be the baseline standard for bicycle and pedestrian data collection. It has a low level of technological sophistication and is labor intensive. Additionally, manual counts allow for data collection of multiple user types, user behaviors and characteristics in addition to volume.

However, automatic counting technology is increasing in use as equipment costs come down and effectiveness of technology improves. Automatic counting technology has higher up-front equipment costs but potential for long-term savings because of reduced labor costs. Additionally, automatic counting technology is useful for long-term counting to establish temporal standards (daily, weekly, yearly volumes and behaviors). Downsides of automatic counting include the need for users to pass a specific point and that the majority of technologies cannot distinguish between bicyclists and pedestrians.

Automatic counting equipment typically falls into one of five categories that include passive infrared, active infrared, video imaging (pixel change analysis or analysis by a person), piezometric (strip or pad), or in-pavement magnetic loop (Jones, 2010, pg. 68). New technology that is being tested for data collection includes GPS devices and GPS-enabled smartphones.

Equipment selection typically includes consideration of a variety of factors such as (Ozbay, 2010, pg. 18):

Commercial availability
Capability to satisfy minimum counting requirements required for needed traffic parameters
Level of technical support provided by vendor
Ease of installation and calibration
Ability to be installed in different positions or heights without impacting performance
Reliability and ability to continuously work in different weather and traffic conditions
Compatibility
Cost-effectiveness of equipment, maintenance and staff labor required for use

Sensor Technologies

Table 1 provides a summary of the technologies found for each of the modes in doing the review. Table 2 breaks out the types of counters in the row labeled “Permanent Counters” as they were named by the authors. Following Table 1 are technology descriptions.

Table 1 Sensor Technologies by Type and Reference

Technology	Pedestrian Counting References	Page	Bicycle Counting References	Page
Impact Sensors/ Tubes	Somasundaram (2010) Ozbay (2010) Jones (2010)	38 18 68	Somasundaram (2010) Jones (2010) Hunter (2009)	38 68
Video	Greene-Roesel (2008) Ismail (2009) Ozbay (2010) Somasundaram (2010)	15 16 18 38	Somasundaram (2010) SRF (2003)	38 39
Permanent Counters	Montufar (2011) Hudson (2010) Somasundaram (2010) Ozbay (2010) Jones (2010) Schneider (2009, “Pilot”) Schneider (2009, “Methodology”) SRF (2003) Greene-Roesel (2008)	17 74 38 18 68 63 62 39 15	Jones (2010) SRF (2010) Nordback (2010)	68 39 27
Portable (Manual)	Jones (2010) Jones (2006) Metropolitan Transportation Commission (2003)	68 48 50	Jones (2010) Jones (2006) Metropolitan Transportation Commission (2003)	68 48 50
Intercept Surveys			Jones (2006) Jones (2010)	48 68
GPS			Casello (2011) Dill (2008) Harvey (2008)	73 30 31
Smartphone Apps			Charlton (2011)	29
Other			Moskovitz (2011) Lovejoy (2011)	81¹ 80²

Table 2 Types of Permanent Counters Discussed

Sensor	Pedestrian Counts	Bicycle Counts
Break-beam with target	Hudson (74)
Computer vision	Green-Roessel (15)	Somasundaram (38)
Electronic piezo		Somasundaram (38)
Hydro acoustic		Somasundaram (38)
Inductive loop	SRF (39)	Nordback (27) Somasundaram (38) SRF (39)
Infrared	Hudson (74) Somasundaram (38)SRF (39)	Somasundaram (38) SRF (39)
Infrared beam	Green-Roessel (15)
Infrared, active	Jones (68)
Infrared, dual	Schneider (63, 62) Hudson (74)
Infrared, passive	Ozbay (18) Jones (68)
Infrared, passive array	Green-Roessel (15)
Infrared, passive dual beam	Green-Roessel (15)
Laser scanner	Green-Roessel (15)
Magnetic loop	Jones (68)	Jones (68)
Mechano acoustic		Somasundaram (38)
Microwave	Montufar (17)
Passive infra-red	Montufar (17)
Peizometric tube	Jones (68)	Jones (68)
Piezo metric pad	Jones (68)	Jones (68)
Piezo-electric pad	Green-Roessel (15)
Pneumatic tube		Somasundaram (38)
Pneumatic/piezo electric/tube counters	SRF (39)	SRF (39)
Pyroelectric	Montufar (17) Somasundaram (38)	Somasundaram (38)
Radio beam	SRF (39)	SRF (39)
Radio beam metal		Somasundaram (38)
Radio beam reflective	Somasundaram (38)	Somasundaram (38)
Stereo-vision curbside detector	Montufar (17)
Thermal	Ozbay (18)	Ozbay (18)
Video imaging	Jones (68)	Jones (68)

Impact Sensors/Tubes

Impact, or piezometric, sensors or tubes use a change in pressure to detect a pedestrian or a person riding a bike. Sensors are fixed at a particular location while tubes can be easily moved. Sensors are more commonly used for pedestrian and bicycle data collection, while tubes are primarily used for bicycle data collection. Impact sensors are best used on a sidewalk or path. Tubes are best used on a path or a street.

Video

Video data collection can be analyzed using computers to detect pixel change or by a person. Computer-based video processing requires significant calibration and is primarily intended for indoor use. However, outdoor applications are being tested with case studies. Manual video processing by a person is effective in collecting volume data as well as user characteristics. It has also been used to analyze the variance of manual counts and other automatic counting equipment because it can be reviewed several times or slowed down to allow a technician time to accurately document data.

Four references address video data collection for pedestrians and two for bikes.

Permanent Counters

Permanent counters require installations at a fixed location and are either not able to be moved once installed or are not easily moved.

For the purpose of this literature review, passive infrared, active infrared and laser scanners are considered permanent pedestrian counters. Passive infrared equipment is best used for counting pedestrian volume along a sidewalk or at intersections. Active infrared can distinguish between bicycles and pedestrians, so it can be used in multiple location including sidewalks and shared paths. Laser scanners can be used along a sidewalk or path but require an open, unobstructed detection area.

For the purpose of this literatures review, active infrared and in-pavement magnetic loop detectors are considered permanent bicycle counters. In-pavement magnetic loop detectors sense a change in magnetic field as metal passes, such as a bike frame, and are appropriate for paths or streets.

Table 2 named twenty-nine technologies including variants that could be considered for permanent count locations. Twenty-four were identified for pedestrian counts and seventeen for bicycle counts. Not all technologies identified were considered to be satisfactory for the specific application tested.

Portable Counters

For the purpose of this literature review, manual pedestrian counts are considered portable counts. Manual counts are a common data collection method. Equipment requirements typically include clip boards, pre-made paper forms, pens or pencils and a hand-held counter. Manual counts are labor intensive as they require a staff person or volunteer to spend time in the field. Standard methods for manual counts have been developed.

Intercept Surveys

Intercept surveys, or surveys of users at a specific site, can be used to collect additional bicycle data beyond volume. Surveys allow data to be collection for user characteristics, behaviors, and preferences.

GPS

GPS devises are typically used to support studies that analyze a sample of a population. GPS devices can be used to collect trip data, such as trip length and duration, and user characteristics, such as demographics and travel behavior. GPS devises are often combined with user surveys to enhance data with additional information about the users.

Smartphone Apps

Smartphone apps can be used to collect trip data, such as route length and duration, and user information, such as demographics and travel behavior. Smartphone apps require the use of a smartphone, development of an app (often for multiple operating systems) and server space for data collection. Since smartphones cannot pick up signals everywhere not all areas can be monitored with this method.

Special Counts

Time series photography can be used to collect data on bicycle parking facilities. Data for volume over time, trip duration and parking behavior was collected as part of a study conducted by Moskovitz (2011, pg. 81).

Another unique source for bicycle data is police reports and hospital records. A study done by Lovejoy (2011, pg. 80) used these sources to collect data on bicycle theft, volume of bicycle injuries and bicycle accidents. A benefit of this data is that it is often standardized, objective and readily available. However, a disadvantage of this data is that incidents are often under-reported.

Pedestrian Counting Equipment

Pedestrian counting equipment needs to be selected based on count location and the purpose of the data collection. If manual counts are used, training and field equipment (such as forms, clip boards, pens, hand counters, etc.) are needed. If automatic count technology is used, field installation, calibration and data retrieval (either remotely or in the field) needs to be considered. The automatic pedestrian counting technology found in the literature includes passive infrared, active infrared, video imaging, piezometric pads and laser scanners.

Information on the use of GPS devices, smartphones and pedometers for pedestrian data collection was not found. Neither was information found on how pedestrian counting equipment can be integrated with motorized vehicle counting equipment. This is particularly important in urban areas where motorized and non-motorized traffic volumes are high and modes are mixed. Additional information is also needed to understand the environmental impacts and constraints on equipment, such as weather, site conditions, traffic modes, traffic volumes, etc.

Fixed Location

Information on fixed location equipment for pedestrian data collection includes impact sensors, video, passive infrared, active infrared, laser scanners and manual counts. Additionally, intercept surveys have been used to supplement data collection at fixed locations.

Among fixed location counting equipment types, video-based technologies are the most commonly cited within the international literature. There are ten citations in this report which used some type of vision technology. There are six discussions of video technology evaluation and four that discuss applications which are listed in the following paragraph.

Video can be augmented by image processing techniques that classify objects and track their paths through the use of algorithms (Li, 2010, pg. 36). Research from Austria (Brandle, 2009, pg. 33) analyzed three different video-based analysis techniques (embedded 3D sensors, motion path modeling, and optical flow) and compared their advantages and disadvantages, finding all three to be largely accurate but also in need of refinement. In Zurich, lasers were used to count pedestrians and map pedestrian movements (Schweizer, 2005, pg. 19). While the counts were accurate and provided useful data in the mapping of movements, off- the-shelf software was not available to interpret the data. EcoCounter (Rheault, 2008, pg. 37) employs acoustic slabs to monitor pedestrian activity. Queensland, Australia employed a matrix of its pedestrian count program needs to determine the most appropriate technology and, perhaps tellingly, constructed a hybrid device taking advantage of multiple components (Davies, 2008, pg. 35).

Citation:

Greene-Roesel, R., M.C. Diogenes, D.R. Ragland. and L.A. Lindau. Effectiveness of a Commercially Available Automated Pedestrian Counting Device in Urban Environments: Comparison with Manual Counts. Presented at the 87th Annual Meeting of the Transportation Research Board 87th Annual Meeting, Washington, D.C., 2008.

Pedestrian Consideration	Addressed?	Bicycle Consideration
Portable Counts	X	Portable Counts
Permanent Counts	X	Permanent Counts
Special Counts	X	Special Counts
Factoring Methods	X	Factoring Methods
Variance of Data	X	Variance of Data
Minimum Counts Needed	X	Minimum Counts Needed
Reporting Information/Customer Needs		Reporting Information/Customer Needs
Utilizing Different Sources	X	Utilizing Different Sources
Formats for Counts	X	Formats for Counts
Reasons for Data Collection	X	Reasons for Data Collection

Summary:

This contains an analysis of the effectiveness of commercially available automatic pedestrian counting devices compared to manual counts. Several automated pedestrian counting devices were considered for use in the analysis, but the dual sensor passive infrared counter was ultimately selected for its cost and ease of deployment. The conclusion was that the device can be used to obtain reasonable estimates of pedestrian volumes in outdoor environments. Additionally, the study found that field observations and manual counts from video-recordings provide relatively accurate counts, but accuracy is dependent on the complexity of the counting task and the level of observer motivation.

Citation:

Ismail, K.A., T.A. Sayed and N. Saunier. Automated Collection of Pedestrian Data Using Computer Vision Techniques. Presented at the 88th Annual Meeting of the Transportation Research Board, Washington, D.C., 2009.

Pedestrian Consideration	Addressed?	Bicycle Consideration
Portable Counts		Portable Counts
Permanent Counts	X	Permanent Counts
Special Counts	X	Special Counts
Factoring Methods	X	Factoring Methods
Variance of Data	X	Variance of Data
Minimum Counts Needed		Minimum Counts Needed
Reporting Information/Customer Needs		Reporting Information/Customer Needs
Utilizing Different Sources		Utilizing Different Sources
Formats for Counts		Formats for Counts
Reasons for Data Collection	X	Reasons for Data Collection

Summary:

This material presents a technique for automated pedestrian data collection using video cameras. Several conclusions are presented including that the accuracy of data is sensitive to camera calibration, environmental conditions can significantly influence data accuracy, and there are limited systematic procedures for evaluating video for pedestrian data collection.

Citation:

Montufar, J. and J. Foord. Field Evaluation of Automatic Pedestrian Detectors in Cold Temperatures. Presented at the 90th Annual Meeting of the Transportation Research Board, Washington, D.C., 2011.

Pedestrian Consideration	Addressed?	Bicycle Consideration
Portable Counts	X	Portable Counts
Permanent Counts	X	Permanent Counts
Special Counts	X	Special Counts
Factoring Methods	X	Factoring Methods
Variance of Data	X	Variance of Data
Minimum Counts Needed	X	Minimum Counts Needed
Reporting Information/Customer Needs		Reporting Information/Customer Needs
Utilizing Different Sources		Utilizing Different Sources
Formats for Counts	X	Formats for Counts
Reasons for Data Collection	X	Reasons for Data Collection

Summary:

This contains the results of conducting a field performance analysis of three commercially available curb-side automatic pedestrian detectors to assess their performance during winter months. A pole-mounted passive infrared and stereovision curb-side detector, a pole-mounted passive infrared curb-side detector and a pole-mounted microwave detector were tested. Initial findings show the passive infrared and microwave detectors perform better at warmer temperatures and the infrared/video detector performs better at colder temperatures.

Citation:

Ozbay, K., B. Bartin, H. Yang, R. Walla and R. Williams. Automatic Pedestrian Counter. Publication FHWA-NJ-2010-001. NJDOT, FHWA, New Jersey Department of Transportation, Federal Highway Administration. 2010.

Pedestrian Consideration	Addressed?	Bicycle Consideration
Portable Counts		Portable Counts
Permanent Counts	X	Permanent Counts
Special Counts		Special Counts
Factoring Methods	X	Factoring Methods
Variance of Data	X	Variance of Data
Minimum Counts Needed		Minimum Counts Needed
Reporting Information/Customer Needs	X	Reporting Information/Customer Needs
Utilizing Different Sources	X	Utilizing Different Sources
Formats for Counts	X	Formats for Counts
Reasons for Data Collection	X	Reasons for Data Collection

Summary:

This is a review of existing pedestrian data collection methods and selection of two commercially available automatic counters to test in the field. The methods of pedestrian data collection reviewed include manual counting as well as the following automatic counting technologies: infrared beam counters, passive infrared counters, piezoelectric pads, laser scanners, and computer vision. The pedestrian counters selected for field testing include a double pyroelectric sensor and a thermal sensor. The counters were selected based on criteria that included the availability, capability, vendor support, ease of deployment, adjustability, reliability, compatibility and cost-effectiveness of a device. To test the devices, the evaluation methodology included test site criteria, data collection methods, and data evaluation procedures. Included are appendices containing interview questions and summaries for professionals currently managing automatic pedestrian counting programs, a summary of case studies of automatic pedestrian counters, general recommendations for an automatic pedestrian counting program, and guidelines for the sensors tested.

Citation:

Schweizer, T. Methods for Counting Pedestrians. Zürich, Switzerland, Presented 6th International Conference on Walking in the 21st Century, Zürich, Switzerland, 2005.

Pedestrian Consideration	Addressed?	Bicycle Consideration
Portable Counts	X	Portable Counts
Permanent Counts		Permanent Counts
Special Counts		Special Counts
Factoring Methods	X	Factoring Methods
Variance of Data	X	Variance of Data
Minimum Counts Needed	X	Minimum Counts Needed
Reporting Information/Customer Needs		Reporting Information/Customer Needs
Utilizing Different Sources		Utilizing Different Sources
Formats for Counts		Formats for Counts
Reasons for Data Collection		Reasons for Data Collection

Summary:

The paper addressed the use of laser detectors for counting pedestrians. The main disadvantage of this method is the lack of off-the-shelf software for interpreting the data. The article did discuss the required number of data points to ensure reasonable accuracy and the potential for extrapolating daily volume from 15-minute or 30-minute counts. Bar and whisker graphs illustrated variances in the data and predicted pedestrian volumes.

Individual-based

No information was found for individual-based equipment for pedestrian data collection in the U.S. However, information on GPS and smartphone apps applied to bicycle data collection was found and may provide cross-over application to pedestrian data collection.

The international literature frequently notes the ability of GPS, smartphone apps and other individual-based technologies to provide detailed information about pedestrian travel routes, speeds, etc. However, such techniques are not applicable to passive counting of pedestrians because they require the subject pedestrians to deploy a particular device in order to be observed.

There is also abundant pedestrian safety research that aims to reduce pedestrian-motor vehicle conflicts through the use of individual vehicle and pedestrian technologies. For example, pedestrians equipped with an electronic tag can send signals to nearby motorists when they enter a particular area such as an intersection that has a magnetic field (Oda, pg. 40). While these examples also currently require the pedestrian to be an active participant, there may be future applications for performing pedestrian counts.

Bicycle Counting Equipment

Bicycle counting equipment needs to be selected based on count location and the purpose of the data collection. If manual counts are used, training and field equipment (such as forms, clip boards, pens, hand counters, etc.) are needed. If automatic count technology is used, field installation, calibration and data retrieval (either remotely or in the field) needs to be considered. The automatic bicycle counting technology found in the literature includes active infrared, video imaging, piezometric, in-pavement magnetic loop, GPS and smartphone apps.

Information was not found on how bicycle counting equipment can be integrated with motorized vehicle counting equipment.

Fixed Location

Information on fixed location equipment for bicycle data collection includes impact sensors, video, active infrared, in-pavement magnetic loop detectors and manual counts. Additionally, intercept surveys have been used to supplement data collection at fixed locations.

Many of the fixed count location equipment types available for the pedestrian mode are also applicable to the bicycle mode, but the research indicates that additional techniques are unique to counting bicyclists such as pneumatic tubes and in-ground inductive loops. There are eleven citations associated with evaluation of fixed location equipment including five in the international literature discussed here.

The City of Hamilton, New Zealand conducted extensive research and testing of a wide variety of bicycle count equipment types (infrared, radar, video, pneumatic tubes, magnetic field detection, and inductive loops) before determining that the in-ground loops would be the most appropriate and cost-effective technology for its own count program (Lieswyn, 2010, pg. 25). Similarly, the Norwegian government tested inductive loop, pneumatic tube, and infrared technologies from multiple companies to determine their respective accuracy levels (Hjelkrem, 2009, pg 24). It is worth noting that while there was some variability amongst the various equipment types and sources, all were found to have an accuracy level of above 83%.

Among more specific tests, a video-based data collection method applicable to both modes, also reports high levels of accuracy (Brandle, 2010, pg. 34). In New Zealand pneumatic tube counters were found to be 100% accurate in an environment exclusive to bicyclists while they were 92% accurate in mixed environments. Speed of the bicyclists was found to be a factor as well as the length of the pneumatic tubes (Macbeth, 2002, pg. 26). VicRoads uses inductive loop counters to obtain 24-hour, 365-day counts at 21 locations in Melbourne, Australia (VicRoads, 2011, pg. 23). Vienna uses radar counters to obtain automated counts (Berger, 2007, pg. 22). These are validated using video surveillance and are reported to be “operating flawlessly.”

Citation:

Berger, Thomas, Workshop: Strategies in a Metropolis Velo Monitoring Vienna, Vienna, Austria. 2007.

Pedestrian Consideration	Bicycle Consideration	Addressed?
Portable Counts	Portable Counts	X
Permanent Counts	Permanent Counts	X
Special Counts	Special Counts	X
Factoring Methods	Factoring Methods	X
Variance of Data	Variance of Data	X
Minimum Counts Needed	Minimum Counts Needed
Reporting Information/Customer Needs	Reporting Information/Customer Needs	X
Utilizing Different Sources	Utilizing Different Sources	X
Formats for Counts	Formats for Counts
Reasons for Data Collection	Reasons for Data Collection	X

Summary:

The City of Vienna (Austria) has a traffic count program which includes manual counts, automated counters, and cyclist interviews. They use radar counters retrofit onto existing poles. Data are checked using video surveillance and counts. The eight counting posts are reported as “operating flawlessly.”

The City also maintains a website http://www.snizek.at/radverkehr/dauerzaehlung2.php on which they publish numerous graphs showing counts by time of day, day of week, and month. Annual summaries and analyses are also provided.

The City is using the data to develop methods for projecting traffic volumes based upon 7-hour counts, correction of traffic due to rainfall, creating daily and seasonal adjustment factors, and identifying trip purpose trends.

Citation:

Cycling Data and Statistics. VicRoads, Melbourne, Australia.http://www.vicroads.vic.gov.au/Home/Moreinfoandservices/Bicycles/StrategicDirectionsForCycling/CyclingDataAndStatistics.htm. Accessed June 17, 2011.

Pedestrian Consideration	Addressed?	Bicycle Consideration	Addressed?
Portable Counts		Portable Counts	X
Permanent Counts		Permanent Counts	X
Special Counts	X	Special Counts	X
Factoring Methods		Factoring Methods
Variance of Data		Variance of Data
Minimum Counts Needed		Minimum Counts Needed
Reporting Information/Customer Needs		Reporting Information/Customer Needs	X
Utilizing Different Sources		Utilizing Different Sources	X
Formats for Counts		Formats for Counts
Reasons for Data Collection		Reasons for Data Collection	X

Summary:

VicRoads has installed permanent inductive loop counters in various locations across Melbourne. Seventeen are off road on pathways and another four are located on road. Each site records cycling flows in two directions. The counters record bicycle volumes 24 hours per day, every day of the year. This is useful for monitoring changes in bicycle use over the seasons but also year to year.

The state government survey “Victorian Integrated Survey of Travel and Activity” (VISTA) took place in 2007 and 2009 and included data on bicycle and pedestrian trips. This data can be accessed at http://www.transport.vic.gov.au/vista.

Local communities also perform manual counts. One regional manual count is held on the first Tuesday in March between 7:00 and 9:00 A.M.

Citation:

Hjelkrem, O.A. and T. Giaever. A Comparative Study of Bicycle Detection Methods and Equipment. Presented at 16th ITS World Congress and Exhibition on Intelligent Transport Systems and Services, Stockholm, Sweden, 2009.

Pedestrian Consideration	Bicycle Consideration	Addressed?
Portable Counts	Portable Counts	X
Permanent Counts	Permanent Counts	X
Special Counts	Special Counts
Factoring Methods	Factoring Methods
Variance of Data	Variance of Data	X
Minimum Counts Needed	Minimum Counts Needed
Reporting Information/Customer Needs	Reporting Information/Customer Needs
Utilizing Different Sources	Utilizing Different Sources	X
Formats for Counts	Formats for Counts	X
Reasons for Data Collection	Reasons for Data Collection	X

Summary:

This documents a study of the accuracy of several types and manufacturers of bicycle counting equipment. The research was undertaken by the Norwegian Public Roads Administration because of reported inadequacies in its current bicycle counting program.

Seven pieces of equipment were tested for accuracy and compared against parallel manual counts: several models of inductive loop equipment (94.0% and 96.4% and one not reported), pneumatic tubes (98.1%), inductive loops for bicycles only (97.5%), inductive loops for bicycles, pedestrians, and motorized vehicles (83.5%), and infrared (84.5%). Numbers of observations among the tests ranged from 100 to several thousand. Common observed reasons for inaccuracies included difficulty separating bicycles passing the equipment at nearly the same time and bicycles passing near the edge of inductive loops.

While not the focus of the research, the paper does include data showing variations in bicycle use by time of day and day of week. In the study location (Trondheim, Norway), counts are much higher on weekdays than on weekends and during the AM and PM peak hours, suggesting that the vast majority of bicycle traffic there is commute-related.

Citation:

Lieswyn, J., A. Wilke and S. Taylor. Automatic Cycle Counting Programme Development in Hamilton. IPENZ Transportation Group Technical Transportation Conference, New Zealand, 2011.

Pedestrian Consideration	Bicycle Consideration	Addressed?
Portable Counts	Portable Counts	X
Permanent Counts	Permanent Counts	X
Special Counts	Special Counts
Factoring Methods	Factoring Methods	X
Variance of Data	Variance of Data	X
Minimum Counts Needed	Minimum Counts Needed	X
Reporting Information/Customer Needs	Reporting Information/Customer Needs	X
Utilizing Different Sources	Utilizing Different Sources	X
Formats for Counts	Formats for Counts	X
Reasons for Data Collection	Reasons for Data Collection	X

Summary:

The City of Hamilton, New Zealand recently introduced a citywide bicycle count program. This paper includes both a summary of existing literature (used to guide the City's decisions) and the process employed to determine the appropriate characteristics of the implemented program.

A need is stated for bicycle count programs to quantify the effectiveness of infrastructure investments, but generally describe a vacuum of reliable guidance on the subject. The literature review uncovered a variety of automatic count types including infrared, ultrasonic, radar, video imaging, piezometric pressure sensitive, and inductive magnetic field loops (the last of which is the most common type employed in North America). Other highlights of the reviewed literature included the importance of control sites, the need for non-manual counts, and the general finding that “there is no standard method of determining the required number and placement of cycle counters for a cycle counting programme.” Regarding appropriate duration, there is evidence that non-permanent counters can be used to indicate yearly volumes. While durations as short as part of a day can be used, a minimum of two week durations for temporary sites are recommended.

Based on the research findings, a network of twelve count sites was established, with two of the locations hosting permanent counts (some existing manual count locations were also retained for calibration purposes). A cost analysis indicated that in-ground inductive loops would be the most effective equipment type for the City's use. Detailed site investigations were conducted to study the technical aspects of implementation and to determine precise counter locations. The Hamilton count program is expected to be fully operational in 2011.

Citation:

Macbeth, A. G. Automatic Bicycle Counting. IPENZ Transportation Group Technical Transportation Conference, New Zealand, 2002.

Pedestrian Consideration	Bicycle Consideration	Addressed?
Portable Counts	Portable Counts	X
Permanent Counts	Permanent Counts
Special Counts	Special Counts
Factoring Methods	Factoring Methods
Variance of Data	Variance of Data	X
Minimum Counts Needed	Minimum Counts Needed
Reporting Information/Customer Needs	Reporting Information/Customer Needs	X
Utilizing Different Sources	Utilizing Different Sources
Formats for Counts	Formats for Counts
Reasons for Data Collection	Reasons for Data Collection

Summary:

This covers an evaluation of the effectiveness and accuracy of two pneumatic tube counters for the purpose of counting bicycles: a bicycle classifier and a vehicle classifier.

The bicycle classifier was found to be 100% accurate in an exclusive bicycle environment and more than 92% accurate in a mixed environment (roadway).The researchers noted it is possible that the classifier was 100% accurate in the roadway environment and the manual counts were inaccurate.

The vehicle classifier was found to be 100% accurate counting bicycles travelling at more than 10 km/h (6 mph) in a mixed environment until tube lengths exceeded 10 meters.

The researchers noted that the vehicle classification systems used in New Zealand do not include exclusive classes for bicycles. Bicycles are either not classified or grouped with motorcycles.

Citation:

Nordback, K. and B.N. Janson. Automated Bicycle Counts: Lessons from Boulder, Colorado. In Transportation Research Record: Journal of the Transportation Research Board, No. 2190, Transportation Research Board of the National Academies, Washington, D.C., 2010, pp. 11-18.

Pedestrian Consideration	Bicycle Consideration	Addressed?
Portable Counts	Portable Counts	X
Permanent Counts	Permanent Counts	X
Special Counts	Special Counts	X
Factoring Methods	Factoring Methods	X
Variance of Data	Variance of Data	X
Minimum Counts Needed	Minimum Counts Needed	X
Reporting Information/Customer Needs	Reporting Information/Customer Needs	X
Utilizing Different Sources	Utilizing Different Sources	X
Formats for Counts	Formats for Counts	X
Reasons for Data Collection	Reasons for Data Collection	X

Summary:

This presents the analysis of the accuracy of inductive loop detectors over a period of several years. The equipment is part of the City of Boulder's (Colorado) permanent count system for its multiuse path network. It was found that inductive loop detectors are proven, low-maintenance, and low-cost technology for bicycle detection. However, the equipment needs to be periodically and properly calibrated, the software needs to be checked for proper setting, and external factors need to be minimized to provide quality, accurate data.

Individual-based

Information on individual-based equipment for bicycle data collection includes GPS and smartphone app technologies. Both allow for rich data collection of user characteristics, user behavior and trip information.

As with the pedestrian mode, GPS-based data is prevalent in the international literature for observing and characterizing bicycle activity (Menghini, 2009, pg. 32). Again, such approaches typically involve active participants and are therefore used for purposes other than pure counting.

Citation:

Charlton, B., J. Hood, E. Sall and M. Schwartz. Bicycle Route Choice Data Collection using GPS-Enabled Smartphones. Presented at the 90^th Annual Meeting of the Transportation Research Board, Washington, D.C., 2011.

Pedestrian Consideration	Bicycle Consideration	Addressed?
Portable Counts	Portable Counts	X
Permanent Counts	Permanent Counts
Special Counts	Special Counts	X
Factoring Methods	Factoring Methods	X
Variance of Data	Variance of Data	X
Minimum Counts Needed	Minimum Counts Needed	X
Reporting Information/Customer Needs	Reporting Information/Customer Needs	X
Utilizing Different Sources	Utilizing Different Sources	X
Formats for Counts	Formats for Counts	X
Reasons for Data Collection	Reasons for Data Collection	X

Summary:

An app was developed for GPS-enabled smartphones to collect bicycle data related to route choice. The use of the smartphone allowed collection of data related to sensitivity to slope, presence of bike lanes and/or bike route destinations, trip purpose and gender. The information collected is being used to inform the San Francisco regional travel model. Other potential applications of this method identified include tracking pedestrians and before and after studies of new bicycle facilities.

Citation:

Dill, J. and J.P. Gliebe. Understanding and Measuring Bicycling Behavior: A Focus on Travel Time and Route Choice. Publication OTREC-RR-08-03. OTREC, Oregon Transportation Research and Education Consortium. 2008.

Pedestrian Consideration	Bicycle Consideration	Addressed?
Portable Counts	Portable Counts	X
Permanent Counts	Permanent Counts
Special Counts	Special Counts	X
Factoring Methods	Factoring Methods
Variance of Data	Variance of Data	X
Minimum Counts Needed	Minimum Counts Needed	X
Reporting Information/Customer Needs	Reporting Information/Customer Needs	X
Utilizing Different Sources	Utilizing Different Sources	X
Formats for Counts	Formats for Counts	X
Reasons for Data Collection	Reasons for Data Collection	X

Summary:

Global positioning system (GPS) technology was used to record trip data for a sample of adults riding their bicycles in Portland, Oregon. A variety of data types were collected including surveys and trip characteristics (environmental conditions, trip length, etc.). Combined with GIS data for infrastructure, data was collected relating to type of cyclist, frequency of bicycling, age, income, overall trips and mileage, trip purposes, trip speeds, time of day and weather, route choice priorities, and routes by facility type. Based on the findings, information was provided that describes how often, why, when and where cyclists ride.

Citation:

Harvey, F., K.J. Krizek and R. Collins. Using GPS Data to Assess Bicycle Commuter Route Choice. Presented at the 87^th Annual Meeting of the Transportation Research Board, Washington, D.C., 2008.

Pedestrian Consideration	Bicycle Consideration	Addressed?
Portable Counts	Portable Counts	X
Permanent Counts	Permanent Counts
Special Counts	Special Counts	X
Factoring Methods	Factoring Methods
Variance of Data	Variance of Data	X
Minimum Counts Needed	Minimum Counts Needed
Reporting Information/Customer Needs	Reporting Information/Customer Needs	X
Utilizing Different Sources	Utilizing Different Sources	X
Formats for Counts	Formats for Counts	X
Reasons for Data Collection	Reasons for Data Collection	X

Summary:

A method to collect bicycle trip data using global positioning system (GPS) technology is presented. The data was used to analyze bicycle route choice using a combination of participant surveys, GPS device data collection, and GIS data. While the data collected provided detailed trip information (distance, time, elevation, route) and behavior characteristics (self-reporting information), challenges were found associated with GPS data, primarily with data processing. After the data was collected, the data required significant filtering and automated scripts to clean the data. However given the challenges, it was still possible to extrapolate bicyclist behavior using the GPS technology.

Citation:

Menghini, G., N. Carrasco, N. Schussler, and K.W. Axhausen. Route Choice of Cyclists in Zurich: GPS-based Discrete Choice Models, Zurich, Switzerland, 2009.

Pedestrian Consideration	Bicycle Consideration	Addressed?
Permanent Counts	Permanent Counts
Special Counts	Special Counts	X
Factoring Methods	Factoring Methods
Variance of Data	Variance of Data
Minimum Counts Needed	Minimum Counts Needed
Reporting Information/Customer Needs	Reporting Information/Customer Needs
Utilizing Different Sources	Utilizing Different Sources
Formats for Counts	Formats for Counts
Reasons for Data Collection	Reasons for Data Collection	X

Summary:

This presents a bicycling route choice model based on data collected in Zurich, Switzerland. The recommended models include variables related to route length, elevation change, and presence of traffic signals (the influence of traffic volume was not examined because of the unavailability of data). An existing database of 11,000 GPS-tracked bicycle trips (extracted from trips made by all modes) made by 2,435 residents was used to identify existing routes. While the research does include a component of bicycle counting in that GPS responders were used to track cyclists, its applicability is limited because it does not relate to passive detection of bicyclists.

Citations addressing both Modes

Fixed Location

Citation:

Brandle, N. Automatic Classification, Counting and Modeling of Non-Motorized Traffic with Video Analytics. Presented at 10th Annual International Conference on Walking and Livable Communities, New York, United States, 2009.

Pedestrian Consideration	Addressed?	Bicycle Consideration	Addressed?
Portable Counts	X	Portable Counts	X
Permanent Counts	X	Permanent Counts	X
Special Counts		Special Counts
Factoring Methods		Factoring Methods
Variance of Data		Variance of Data
Minimum Counts Needed		Minimum Counts Needed
Reporting Information/Customer Needs		Reporting Information/Customer Needs
Utilizing Different Sources		Utilizing Different Sources
Formats for Counts		Formats for Counts
Reasons for Data Collection	X	Reasons for Data Collection	X

Summary:

This presentation highlights the Austrian Institute of Technology's experience in bicycle and pedestrian monitoring using video analysis. Three video-based analysis techniques are described: 1) an embedded visual 3D sensor, 2) path modeling based on traveler's motion paths, and 3) pedestrian path modeling based on optical flow. The first type has the advantage of real-time counting and has high reported accuracy rates. Individual path modeling is reported to break down in dense bicycle/pedestrian environments because of occlusion. That problem is partially addressed through optical flow, which models crowds as a set of interacting particles.

Citation:

Brandle, N., A.N. Belbachir, and S. Schraml. SmartCountplus – Towards Automated Counting and Modeling of Non-Motorised Traffic with a Stand-Alone Sensor Device. In REAL CORP 2010 Proceedings, Vienna, Austria, 2010, pp. 1261-1266.

Pedestrian Consideration	Addressed?	Bicycle Consideration	Addressed?
Portable Counts	X	Portable Counts	X
Permanent Counts	X	Permanent Counts	X
Special Counts		Special Counts
Factoring Methods		Factoring Methods
Variance of Data		Variance of Data
Minimum Counts Needed		Minimum Counts Needed
Reporting Information/Customer Needs		Reporting Information/Customer Needs
Utilizing Different Sources		Utilizing Different Sources
Formats for Counts		Formats for Counts
Reasons for Data Collection	X	Reasons for Data Collection	X

Summary:

The authors describe a new device used for automated counts of bicyclists and pedestrians. It is a visual sensor that uses changing light intensity to determine depth and therefore object shapes. The key feature of the device is that it is designed to discriminate between pedestrians and bicyclists (and their sub-categories) in an outdoor environment while avoiding privacy issues through the use of “captured depth data.” The device uses embedded clustering (grouping together objects associated with the same traveler) and classification (by mode) algorithms. At a test site with a relatively small sample of 128 facility users, there was a correct classification rate of 92% for riding cyclists and 100% for traditional pedestrians. Walking cyclists and pedestrians with umbrellas were separately analyzed; the latter group had a lower successful classification rate, suggesting the need for further investigation.

Citation:

Davies, R. Pedestrian and Cycling Counters – South East Queensland. Presented at 9th Annual International Conference on Walking and Livable Communities, Barcelona, Spain, 2008.

Pedestrian Consideration	Addressed?	Bicycle Consideration	Addressed?
Portable Counts	X	Portable Counts	X
Permanent Counts	X	Permanent Counts	X
Special Counts	X	Special Counts	X
Factoring Methods		Factoring Methods
Variance of Data		Variance of Data
Minimum Counts Needed		Minimum Counts Needed
Reporting Information/Customer Needs		Reporting Information/Customer Needs
Utilizing Different Sources	X	Utilizing Different Sources	X
Formats for Counts	X	Formats for Counts	X
Reasons for Data Collection	X	Reasons for Data Collection	X

Summary:

The Active Transport Planning Unit of Queensland (Australia) Transport's Integrated Transport Planning Division sought improved bicycle and pedestrian counting techniques and recommendations to better gauge network performance and guide investment decisions. Noting the numerous count technologies, Queensland developed a set of requirements based on its needs, including the ability to count both modes directionally, producing 24-hour counts at 15-minute intervals, allowance of remote access data retrieval, and not being cost prohibitive. After employing a matrix of technologies and needs, the agency ultimately created and implemented its own prototype. The presentation concludes with a discussion of implementation at one site and the daily variation of bicycle and pedestrian volumes at that location.

Citation:

Li, J., C. Shao, W. Xu, and J. Li. Real-time System for Tracking and Classification of Pedestrians and Bicycles. In Transportation Research Record: Journal of the Transportation Research Board, No. 2198, Transportation Research Board of the National Academies, Washington, D.C., 2010, pp. 83-92.

Pedestrian Consideration	Addressed?	Bicycle Consideration	Addressed?
Portable Counts	X	Portable Counts	X
Permanent Counts	X	Permanent Counts	X
Special Counts		Special Counts
Factoring Methods		Factoring Methods
Variance of Data		Variance of Data
Minimum Counts Needed		Minimum Counts Needed
Reporting Information/Customer Needs		Reporting Information/Customer Needs
Utilizing Different Sources		Utilizing Different Sources
Formats for Counts		Formats for Counts
Reasons for Data Collection	X	Reasons for Data Collection	X

Summary:

This is an evaluation of a pedestrian and bicycle tracking and classification (PBTC) system. The system uses typical video-based data collection and then incorporates image-processing technology to detect, track, and classify pedestrians and bicyclists. The algorithm approach of the PBTC system includes six modules.

Three test sites in Beijing, China were examined. An accuracy level of approximately 85% for both modes at all three test sites was observed. Inaccuracies were largely observed to be the result of inaccurate detection (such as double counting pedestrians who stop for long periods of time) and occlusion failure (including when two closely spaced pedestrians are observed as one).

The conclusion was that the PBTC system is generally effective, but that the embedded algorithms could be improved with further research.

Citation:

Rheault, J. Measuring Walking: Counting Pedestrians. Presented at 9th Annual International Conference on Walking and Livable Communities, Barcelona, Spain, 2008.

Pedestrian Consideration	Addressed?	Bicycle Consideration	Addressed?
Portable Counts	X	Portable Counts	X
Permanent Counts	X	Permanent Counts	X
Special Counts		Special Counts
Factoring Methods		Factoring Methods
Variance of Data		Variance of Data
Minimum Counts Needed		Minimum Counts Needed
Reporting Information/Customer Needs	X	Reporting Information/Customer Needs	X
Utilizing Different Sources		Utilizing Different Sources
Formats for Counts		Formats for Counts
Reasons for Data Collection	X	Reasons for Data Collection

Summary:

This presentation was delivered by vendor employee and describes the types and uses of the company's products, which include inductive loops and selective pneumatic tubes for bikes, acoustic slabs for pedestrians, and pyroelectric sensors for both modes. Other components of the presentation include key features, an example output showing volumes by time of day, maintenance recommendations, and data management options.

Citation:

Somasundaram, G., V. Morellas and N.P. Papanikolopoulos. Practical Methods for Analyzing Pedestrian and Bicycle Use of a Transportation Facility. Publication MN/RC 2010-06. MnDOT, Minnesota Department of Transportation. 2010.

Pedestrian Consideration	Addressed?	Bicycle Consideration	Addressed?
Portable Counts		Portable Counts
Permanent Counts	X	Permanent Counts	X
Special Counts		Special Counts	X
Factoring Methods	X	Factoring Methods	X
Variance of Data	X	Variance of Data	X
Minimum Counts Needed		Minimum Counts Needed
Reporting Information/Customer Needs	X	Reporting Information/Customer Needs	X
Utilizing Different Sources	X	Utilizing Different Sources	X
Formats for Counts	X	Formats for Counts	X
Reasons for Data Collection	X	Reasons for Data Collection	X

Summary:

This document contains a review of existing technologies used to collect automatic counts for bicycles and pedestrians, and selection of a video-based system to analyze its application in the field. Several computer vision algorithms were reviewed and tested in the field to assess effectiveness in collection data. The existing technologies reviewed include buried pressure pads, pyroelectric sensor, inductive loop, pneumatic/piezo electric sensors/tube counters, radio beam sensors, infrared sensors, and vision-based sensing. The appendix provides a useful summary table for these methods and summarizes device type, ease of use, type of data collection (bicycle, pedestrian or both), cost, light conditions, accuracy, environmental sensitivity, applicable sites, whether the technology is wireless, power source, and portability. Additionally several algorithms were analyzed in the field to test different conditions and applications of the video-based system.

Citation:

SRF Consulting Group, Inc. Bicycle and Pedestrian Detection. Minnesota Department of Transportation, 2003.

Pedestrian Consideration	Addressed?	Bicycle Consideration	Addressed?
Portable Counts	X	Portable Counts	X
Permanent Counts	X	Permanent Counts	X
Special Counts		Special Counts
Factoring Methods		Factoring Methods
Variance of Data	X	Variance of Data	X
Minimum Counts Needed		Minimum Counts Needed
Reporting Information/Customer Needs	X	Reporting Information/Customer Needs	X
Utilizing Different Sources	X	Utilizing Different Sources	X
Formats for Counts	X	Formats for Counts	X
Reasons for Data Collection	X	Reasons for Data Collection	X

Summary:

This is an early review of the application of sensor technology for non-motorized traffic detection, how the sensor technology is applied in the field and how to conduct field tests to evaluate the performance of the technology. Field applications for the sensor technology reviewed include curbside pedestrian detection, crosswalk pedestrian detection, intersection approach bicycle detection and historical data collection (volume, speed and classification data). Technology reviewed includes commercially available passive infrared/ultrasonic, infrared, microwave, video and magnetic devices. The field tests include documentation of sensor installation, the data collection systems, baseline data requirements, and data collection. The conclusion was that the type of sensor technology should be selected based on the type of data needed, such as real-time data versus historical data. The appropriate applications for each technology reviewed are also summarized.

Individual Based

Citation:

Oda, H., S. Kubota, and Y. Okamoto. Movement-Pattern Models of Pedestrians at Intersections and the Technology to Reduce Traffic Accidents Involving Pedestrians or Cyclists, Yokosuka, Japan.Undated.

Pedestrian Consideration	Addressed?	Bicycle Consideration	Addressed?
Permanent Counts		Permanent Counts
Special Counts		Special Counts
Factoring Methods		Factoring Methods
Variance of Data		Variance of Data
Minimum Counts Needed		Minimum Counts Needed
Reporting Information/Customer Needs		Reporting Information/Customer Needs
Utilizing Different Sources	X	Utilizing Different Sources	X
Formats for Counts		Formats for Counts
Reasons for Data Collection	X	Reasons for Data Collection	X

Summary:

The intent of this research is to refine a system by which motorists are alerted to the presence of pedestrians and bicyclists at intersections. When pedestrians and bicyclists equipped with an electronic tag enter a magnetic field, a signal is sent to motorists (whose vehicles also have a receiver) who may be in a position to conflict with the non-motorized user. The authors describe a matrix-based approach to determining the speed and direction of pedestrians and bicyclists.

The research is aimed at reducing crashes, and because users are required to be equipped with a device, the system is not well-suited for passive bicycle/pedestrian counts.

DATA COLLECTION PROGRAMS

Data collection programs cover any type of program that performs a regularly scheduled survey or volume collection activity. Regularly scheduled surveys as the label implies do not necessarily use equipment except as an adjunct to the individual collecting the data. More emphasis has been placed on volume collection programs that do not require on-site manual activities to generate volumes. Both are included as input to creation of possible standard record formats for storing and sharing data.

A case study of state, regional and local agencies was conducted in 2005 to identify bicycle and pedestrian data collection trends in the United States. The results of the study identified several reasons for having a program including (Schneider, 2005, pg. 44):

Documenting conditions and conducting trend analysis
Bicycle and pedestrian network planning
Pedestrian vehicle crash analysis
Forecasting demand for walking trips
Pedestrian facility design
Traffic law enforcement impact studies
Policy and program design

In addition to the reasons for having a bicycle and pedestrian data collection program, several reasons for not having a program – or a more robust program - were identified as well and include (Schneider, 2005, pg. 44):

Lack of knowledge on how pedestrian data can be integrated with motor vehicle data
Costs for counting equipment and labor for processing and managing data
Funding frameworks for programs
Lack of guidance on pedestrian behavior and demand characteristics
Lack of guidance on data application approaches
Lack of guidance on data gathering procedures
Lack of understanding of the data types needed
Data management
Lack of dissemination standards or platforms for sharing data with users

Table 3 contains the materials identified that specifically reference design or surveys of data collection programs. Programs mentioned in the literature reviewed as a data source are also mentioned in this chapter.

Table 3 Information on Data Programs by Jurisdictional Level

Organization	Pedestrian	Page	Bicycle	Page
National	Jones (2006)	48	Jones (2006) Richardson (2006)	48 54
State	Schneider (2005) Cottrell (2003)	44 46	Schneider (2005)	44
Regional	Schneider (2005) Cottrell (2003) Metropolitan Transportation Commission (2003)	44 46 50	Schneider (2005) Metropolitan Transportation Commission (2003)	44 50
County	Schneider (2005) Cottrell (2003)	44 46	Schneider (2005)	44
City	Schneider (2005) Cottrell (2003) Little (2008) VicRoads (2011)	44 46 52 23	Schneider (2005)	44

Citation:

Bicycle andPedestrian Data: Sources, Needs and Gaps. Publication BTS00-02. US DOT, BTS, U.S. Department of Transportation, Bureau of Transportation Statistics, 2000.

Pedestrian Consideration	Addressed?	Bicycle Consideration	Addressed?
Portable Counts		Portable Counts
Permanent Counts		Permanent Counts
Special Counts		Special Counts
Factoring Methods		Factoring Methods
Variance of Data		Variance of Data
Minimum Counts Needed		Minimum Counts Needed
Reporting Information/Customer Needs	X	Reporting Information/Customer Needs	X
Utilizing Different Sources	X	Utilizing Different Sources	X
Formats for Counts	X	Formats for Counts	X
Reasons for Data Collection	X	Reasons for Data Collection	X

Summary:

Included (circa 2000) are existing data sources, data needs, and options for improving bicycle and pedestrian data. Related to counting, there is an outline of the then current state of practice for counting data collection, how different counting data is used for analysis, and the priorities for data improvement. Recommendations are provided on how to improve data collection programs and standards.

Citation:

Schneider, R.J., R.S. Patten and J.L. Toole. Case Study Analysis of Pedestrian and Bicycle Data Collection in U.S. Communities. In Transportation Research Record: Journal of the Transportation Research Board, No. 1939, Transportation Research Board of the National Academies, Washington, D.C., 2005, pp. 77-90.

Pedestrian Consideration	Addressed?	Bicycle Consideration	Addressed?
Portable Counts	X	Portable Counts	X
Permanent Counts	X	Permanent Counts	X
Special Counts		Special Counts
Factoring Methods		Factoring Methods
Variance of Data	X	Variance of Data	X
Minimum Counts Needed		Minimum Counts Needed
Reporting Information/Customer Needs	X	Reporting Information/Customer Needs	X
Utilizing Different Sources	X	Utilizing Different Sources	X
Formats for Counts	X	Formats for Counts	X
Reasons for Data Collection	X	Reasons for Data Collection	X

Summary:

Twenty-nine communities were surveyed to profile bicycle and pedestrian data collection programs and how they are managed. They also identified the geographic size of agencies collecting data, the scale of data collection programs, data collection methods employed by the different agencies, and frameworks for bicycle and pedestrian data collection programs. A variety of different agencies were surveyed include cities, counties, regional councils, metropolitan planning organizations and state agencies. Methods of data collection varied and included surveys of users, manual counts, automatic counts, and facilities inventories. Data program elements identified include benefits of pedestrian and bicycle data collection, reasons communities do not collect pedestrian and bicycle data, efficiency of data collection processes, scope of data collection efforts, data collection processes, and institutionalization of data collection programs.

PEDESTRIAN PROGRAMS

Pedestrian data collection can be found at all levels of government, although data collection is tailored to local needs (Cottrell, 2003, pg. 46). Programs typically include at least one, and most often several, of the following components:

Manual counting, automatic counting, or both
Surveys identifying user characteristics, behaviors and preferences
Inventories of facilities
Spatial analysis and documentation, often using CAD or GIS
Data collection methods
Data storage
Institutionalization of program

Gaps in the research scanned include a lack of information on staffing organization and responsibilities, how programs fit within the larger agency structure, and how programs work with other agencies to share information. Additionally, there is a need for a set of uniform, national formats for non-motorized transportation data. The uniform formats would allow for local data to be aggregated up to regional, state and national levels, compare data on use and facilities between geographies, and establish, measure, and monitor benchmarks for non-motorized travel and facilities (Schneider, 2005, pg. 44).

Citation:

Cottrell, W. D. and D. Pal. Evaluation of pedestrian data needs and collection efforts. In Transportation Research Record: Journal of the Transportation Research Board, No. 1828, Transportation Research Board of the National Academies, Washington, D.C., 2003, pp. 12-19.

Pedestrian Consideration	Addressed?	Bicycle Consideration
Portable Counts		Portable Counts
Permanent Counts	X	Permanent Counts
Special Counts		Special Counts
Factoring Methods		Factoring Methods
Variance of Data		Variance of Data
Minimum Counts Needed		Minimum Counts Needed
Reporting Information/Customer Needs	X	Reporting Information/Customer Needs
Utilizing Different Sources	X	Utilizing Different Sources
Formats for Counts	X	Formats for Counts
Reasons for Data Collection	X	Reasons for Data Collection

Summary:

An early 2000s survey of several state, regional, and local transportation agencies to identify whether they have a pedestrian data collection program and, if they do, analyzed the structure of the program. The survey identified several program elements including average count duration, count duration range, most common counting times, counting methods and purpose of counts. The three common data collection methods include manual counting, recording push button use and video camera recording. Based on the findings, a pedestrian data collection framework was presented that can be used by local, regional and state planning agencies. The framework outlines the type of data to be collected and how the data can be used to address specific traffic design and planning applications.

While there are examples of pedestrian data collection programs, the quality and quantity of data is typically identified as variable and in need of improvement. Areas of improvement and need for data quality and quantity include (Bicycle and Pedestrian Data, 2000, pg. 43):

Data collection needs to address usage patterns of individual facilities as well as total usage for geographic areas, such as for pedestrian trips for an entire city
Data needs to be collected on a systematic and regular basis to compare patterns over time
Data needs to be collected in consistent manner so that it can be aggregated from local location to large geographic areas, such as a city or region
Data needs to be categorized by type to include information such as user characteristics, trip purpose and length, time of day, and type of facility.

National

The National Bicycle and Pedestrian Documentation Project is a voluntary data collection project that was created through a partnership between Alta Planning + Design and the Institute of Transportation Engineers Pedestrian and Bicycle Council. The goals of the project are to establish a consistent national bicycle and pedestrian count and survey methodology, establish a national database of bicycle and pedestrian count information generated by these consistent methods and practices, and to use the county and survey information to begin the analysis on the correlations between local demographics, climate and land-use factors and bicycle and pedestrian activity.

Citation:

Jones, M.G. and A.M. Cheng. National Bicycle And Pedestrian Documentation Project. Presented at 85th Annual Meeting of the Transportation Research Board, Washington, D.C., 2006.

Pedestrian Consideration	Addressed?	Bicycle Consideration	Addressed?
Portable Counts	X	Portable Counts	X
Permanent Counts	X	Permanent Counts	X
Special Counts	X	Special Counts	X
Factoring Methods	X	Factoring Methods	X
Variance of Data	X	Variance of Data	X
Minimum Counts Needed	X	Minimum Counts Needed	X
Reporting Information/Customer Needs	X	Reporting Information/Customer Needs	X
Utilizing Different Sources	X	Utilizing Different Sources	X
Formats for Counts	X	Formats for Counts	X
Reasons for Data Collection	X	Reasons for Data Collection	X

Summary:

This is a presentation of a preliminary framework for national data collection for bicycle and pedestrian counting. The structure of the project addresses the following: establish consistent survey methodology, establish national database of bicycle and pedestrian count information and use information collected to analyze social, environmental and facilities conditions that influence bicycle and pedestrian activity. The paper provides background information on the project, objectives, methodology and next steps.

Statewide

The information presented by Schneider (2005, pg. 44) documents state level programs as case studies, but no literature was found profiling state-level pedestrian data collection programs. Additionally, a comprehensive list of state data collection programs was not found during the literature review.

For information on how local research can be adopted at the state level, see resource by Jones (2010, pg. 68).

A request for information from State Bicycle and Pedestrian Coordinators found that Washington State has implemented a Washington State Bicycle and Pedestrian Documentation Project in response to the National Documentation Project. This project uses the same methodology as the national project. Rhode Island has installed trail user counters at some of its trail crossings; they will not differentiate between bicyclists and pedestrians.

Regional

The Metropolitan Transportation Commission (MTC), which is the San Francisco region's metropolitan planning organization, created the Bicyclist and Pedestrian Data Collection and Analysis Project to provide a bicycle and pedestrian data collection standard for the local jurisdiction in the region and to provide a central collection point for data collected by local jurisdictions. As part of the project, the MTC developed the Handbook for Bicycle and Pedestrian Counts to establish data collection standards and procedures.

Citation:

Handbook for Bicyclist and Pedestrian Counts. Metropolitan Transportation Commission, (San Francisco, California) 2003.

Pedestrian Consideration	Addressed?	Bicycle Consideration	Addressed?
Portable Counts	X	Portable Counts	X
Permanent Counts	X	Permanent Counts
Special Counts		Special Counts	X
Factoring Methods		Factoring Methods
Variance of Data	X	Variance of Data	X
Minimum Counts Needed	X	Minimum Counts Needed	X
Reporting Information/Customer Needs	X	Reporting Information/Customer Needs	X
Utilizing Different Sources	X	Utilizing Different Sources	X
Formats for Counts	X	Formats for Counts	X
Reasons for Data Collection	X	Reasons for Data Collection	X

Summary:

The handbook was developed to establish bicycle and pedestrian count standards for the Metropolitan Transportation Commission's (MTC) Bicyclist and Pedestrian Data Collection and Analysis Project. The five sections of the handbook address development of a counting calendar, staffing, equipment and forms standards, schedule of activities, data processing and management, and user survey procedures. The section that addresses development of the count calendar includes information on when counts should be conducted, summary of factors that may influence the counts, and the creation of a master count schedule. The section that addresses staffing, equipment and forms addresses staff requirements and their roles, outlines equipment options, and provides standardized forms with descriptions on how to use them. The section on schedule of activities addresses pre-planning and day of requirements. The section on data processing and management outlines how the field data should be transferred to a database and how the database should be managed. The final section on user surveys outlines procedures for administration and analysis of user surveys.

Local/City/Municipal

The information presented by Schneider (2005, pg. 44) documents local-level programs as case studies, but no literature was found profiling specific local-level pedestrian data collection programs in the U.S.

For information on how research and a data collection program can be applied at the local level, see resource by Jones (2010, pg. 68).

London, England expressed a need to better understand the extent of walking in the city (Little, 2008, pg. 52). Accordingly, the city's transportation agency tested numerous technologies to determine accuracy of pedestrian counts. Based on the findings, the city has now installed 40 video-based count locations.

Melbourne, Australia includes pedestrian trip questions within its “Victorian Integrated Survey of Travel and Activity” (VicRoads, 2011, pg. 23). The survey includes walking as a potential travel mode with respect to trip chaining and travel group size. While it asks questions about specific routes chosen for vehicular trips, it does not do so for pedestrian trips.

Citation:

Little, B. Pedestrian Monitoring. Presented at 9th Annual International Conference on Walking and Livable Communities, Barcelona, Spain, 2008.

Pedestrian Consideration	Addressed?	Bicycle Consideration
Portable Counts	X	Portable Counts
Permanent Counts	X	Permanent Counts
Special Counts		Special Counts
Factoring Methods		Factoring Methods
Variance of Data	X	Variance of Data
Minimum Counts Needed		Minimum Counts Needed
Reporting Information/Customer Needs	X	Reporting Information/Customer Needs
Utilizing Different Sources	X	Utilizing Different Sources
Formats for Counts		Formats for Counts
Reasons for Data Collection	X	Reasons for Data Collection

Summary:

This presentation documents the efforts of London Streets, a City of London transportation agency, to decide upon and implement a pedestrian counting/monitoring program. The City's mayor had stated that the true extent and nature of walking in London is poorly understood and stated an aim to monitor the extent of walking in order to set a baseline for walking in London.

Site selection was determined on the basis of popular destinations, cordon or screenlines, key routes, and random locations. Trials were conducted for several identified technologies including closed circuit television detectors, floor mounted laser scanners, passive infrared array detectors, and overhead laser ranging scanners. The specific platforms examined included laser, video and CCTV. Manual counts were compared against the automated counts, with error percentages ranging from 1.6% (video) to 8.1% (CCTV). On the basis of the trials, video option was awarded a three-year contract for 40 monitoring sites.

The presentation concludes by highlighting several example monitoring sites by showing a photograph of the installation and identifying associated land use, location, required permission, mounting type, processor location, and power source.

BICYCLE PROGRAMS

Similar to pedestrian data collection programs, bicycle data collection programs can be found at all levels of government. While some agencies have developed their own programs, most agencies have used the National Bicycle and Pedestrian Documentation Project to advance the systematic collection of data for bicycle use.

Additionally, data vary by geography and government agency. Some agencies have robust data collection programs that are easily applied across geographies and compared to other data collected by different agencies, some data is collected sporadically to address specific local needs and some agencies do not collect data at all. The lack of consist quantities and types of data make it difficult for data users to compare data to geographies outside of the location where the data is collected (Jones, 2010, pg. 68).

National

Outside the United States, at least one other nation has embarked on a nationwide counting project. Switzerland maintains a national bicycle route system, Veloland Schweiz, and the Swiss government undertook a study to determine the system's usage and associated economic benefits (Richardson, 2006). Because of the relatively small geographic scale of the network, counts (and associated surveying of intercepted bicyclists) were able to be performed manually. One of the most instructive components of the Swiss count program is that it included the development of count weighting factors based on weather and day of week, along with spatial weights that can be used to extrapolate point-based counts to an entire bicycle network.

Citation:

Richardson, A.J. Estimating Bicycle Usage on a National Cycle Network. In Transportation Research Record: Journal of the Transportation Research Board, No. 1982, Transportation Research Board of the National Academies, Washington, D.C., 2006, pp. 166-173.

Pedestrian Consideration	Bicycle Consideration	Addressed?
Portable Counts	Portable Counts	X
Permanent Counts	Permanent Counts
Special Counts	Special Counts
Factoring Methods	Factoring Methods	X
Variance of Data	Variance of Data
Minimum Counts Needed	Minimum Counts Needed
Reporting Information/Customer Needs	Reporting Information/Customer Needs	X
Utilizing Different Sources	Utilizing Different Sources
Formats for Counts	Formats for Counts	X
Reasons for Data Collection	Reasons for Data Collection	X

Summary:

This describes a multi-year survey of bicyclists using Veloland Schweiz, a national cycling route system. The survey included basic counts as well as follow-up interviews and mail-back questionnaires. The purpose of the survey was to estimate network-wide usage and the associated bicycling-related expenditures made by users of the system.

The counts themselves were performed manually – one or two people hand-counted bicyclists as they passed each of 16 survey sites and noted basic characteristics of the cyclists. A random sample of those counted cyclists was chosen to participate in the follow-up interview.

To convert the sample counts to a total estimate of system users, several weighting factors were applied: nonresponse, nonacceptance, cycle flow, site selection, temporal expansion, and spatial expansion. Two temporal factors were considered, seasonal (weather) weights and day-of-week weights. For the former, the temperature and rainfall were recorded for each of the sixteen survey dates. These measurements were compared against cycle flow rate graphs that show the propensity of cycling (broken into several trip characteristics) by temperature and rainfall, and then scaled accordingly. The graphs, based on prior research, suggest that cycling is most prevalent at a temperature of approximately 25° C (approximately 77° F) with no rainfall. The surveys were conducted on Wednesdays and Sundays, so the day-of-week adjustments simply make the assumption that Wednesday counts are representative of other weekdays and Sunday counts are representative of Saturday as well.

Local/City/Municipal

Internationally, several municipal governments have established methods for monitoring bicycle activity. The city of Vienna, Austria uses radar counters and video surveillance to track bicycling activity and has methods for projecting short-term counts (Berger, 2007, pg. 22). Hamilton, New Zealand instituted a bicycle counting program following its own extensive review of the literature and available counting techniques (Lieswyn, 2010, pg. 25). The recently implemented program largely uses in-ground inductive loops and includes a mix of permanent and temporary count stations. Melbourne, Australia's VicRoads has installed and monitors permanent count stations at 21 locations across the city (VicRoads, 2011, pg. 23). Their website includes comprehensive reports analyzing hourly, daily, and monthly variations in bicycle traffic. The communities within Melbourne supplement these automated counts with a one-day 2-hour manual count each March.

DATA PROCESSING AND STORAGE

Data processing and storage addresses how data is handled after it is collected and what methods are used to support future retrieval.

Table 4 Sources with Information on Factoring, Sampling and Count Variance

Source	Page	Pedestrian			Bicycle
Source	Page	Factoring	Count Variance	Sampling	Factoring	Count Variance	Sampling
Berkow (2009)	72				X	X	X
Casello (2011)	73				X	X	X
Charlton (2011)	29				X	X	X
Dill (2008)	30				X	X	X
Diogenes (2007)	60	X	X
Greene-Roesel (2008)	15	X	X
Harvey (2008)	31				X	X	X
Hudson (2010)	74	X	X	X	X	X	X
Hunter (2010)	76				X	X	X
Ismail (2009)	16	X		X
Jones (2006)	48	X	?	?	X	X	?
Jones (2010)	68	X	X	X	X	X	X
Lovejoy (2011)	80				X	X	X
Metropolitan Transportation Commission (2003)	50			X
Montufar (2011)	17	X	X	X
Moskovitz (2011)	81				X	X	X
Nordback (2010)	27				X	X
Ozbay (2010)	18	X	X	X
Schneider (2009, “Methodology”)	62	X	X	X
Schneider (2009, “Pilot”)	63	X	X	X
Somasundaram (2010)	38	X			X	X	X
SRF (2003)	39		X		X	X	X

PEDESTRIAN COUNTS

Information on pedestrian data processing and storage is largely found through case studies. The information for data processing and storage in these case studies generally falls into categories of factoring methods, count variance, or sampling

The international literature search did not reveal significant information on pedestrian sampling techniques, recording formats, or file structures. While researchers and practitioners whose work was reviewed for this literature search must have developed recording formats and file structures for their analyses, these were not discussed in any detail in the papers reviewed. Presumably, further exploration and follow-up correspondence with the authors would allow for the collection of format and structure information.

The major gaps in the literature are related to data formats and file structures. While additional information is needed for factoring methods, count variance, and sampling, there is very little information in the literature review that addresses record formats and file structures in detail.

Factoring Methods

Factoring methods help reduce error and variances in data. Additionally, factoring methods are used to extrapolate findings from sites to similar locations, aggregated geographies or generalize data for modeling purposes. Factoring methods typically use a baseline data set, a sample data set of field data and statistical methods to process data. Manual counts, video processing by a person, or seasonal counts are typically used to establish a baseline data set or factoring measure.

Frequently, limitations in budget or equipment lead to the conducting of temporary, short-term counts even though agencies are interested in long-term pedestrian monitoring. As a result, significant portions of the international literature are devoted to methods that can be used to factor counts over longer periods of time. Two frequently discussed factoring methods involve day-of-week and seasonal variations.

Day-of-week variability in activity depends to a significant extent on the nature of that activity (i.e. trip purpose) (Mondheim, 1998, pg. 61). In areas where pedestrian travel is predominantly utilitarian in nature, including commute trips, activity is higher on weekdays. Where recreational activity is more prevalent, the converse may be true.

Count Variance

All pedestrian data collection has some degree of error or bias. Count variance can be attributed to several factors including equipment technology, environmental factors, site conditions, seasonal, weekly or time of day factors, and human error.

Sampling

Sampling for pedestrian data collection can be applied to site selection, day and time, or participant selection for studies that track individuals. Methods for site selection range from the use of local knowledge of facilities, facility characteristic, facility volumes, geography or spatial characteristics, such as urban or rural. Selection criteria range from the use of site conditions criteria to statistical methods. Day and time selections are often made to capture peak periods for pedestrian volumes and to capture seasonal or weekday variations.

Citation:

Cullen, P. Effective Pedestrian Surveys with Everyday Equipment. Presented at8th Annual International Conference on Walking and Livable Communities, Toronto, Canada, 2007.

Pedestrian Consideration	Addressed?	Bicycle Consideration
Portable Counts	X	Portable Counts
Permanent Counts		Permanent Counts
Special Counts	X	Special Counts
Factoring Methods		Factoring Methods
Variance of Data	X	Variance of Data
Minimum Counts Needed		Minimum Counts Needed
Reporting Information/Customer Needs	X	Reporting Information/Customer Needs
Utilizing Different Sources		Utilizing Different Sources
Formats for Counts	X	Formats for Counts
Reasons for Data Collection	X	Reasons for Data Collection

Summary:

In this conference “note,” the author describes the highly variable nature of pedestrian trips and behaviors and makes the case for improved surveying techniques. With regard to measuring pedestrian flows, it is suggested that time increments need to be very short (one minute) to account for significant minute-by-minute differences. Most of the pedestrian survey techniques discussed involve self-survey using Personal Data Assistants, cell phone applications, and other personal data loggers.

Citation:

Diogenes, M. C., R. Greene-Roesel, L.S. Arnold and D.R. Ragland. Pedestrian Counting Methods at Intersections: A Comparative Study. In Transportation Research Record: Journal of the Transportation Research Board, No. 2002, Transportation Research Board of the National Academies, Washington, D.C., 2007, pp. 26-30.

Pedestrian Consideration	Addressed?	Bicycle Consideration
Portable Counts	X	Portable Counts
Permanent Counts	X	Permanent Counts
Special Counts		Special Counts
Factoring Methods	X	Factoring Methods
Variance of Data	X	Variance of Data
Minimum Counts Needed		Minimum Counts Needed
Reporting Information/Customer Needs	X	Reporting Information/Customer Needs
Utilizing Different Sources	X	Utilizing Different Sources
Formats for Counts	X	Formats for Counts
Reasons for Data Collection	X	Reasons for Data Collection

Summary:

This is an analysis of the effectiveness and accuracy of three pedestrian counting methods including manual counting with sheets, manual counting with a clicker and manual counting with video. The results of the analysis show pedestrian counts taken in the field were systematically lower than the pedestrian counts taken by observing video recordings. The accuracy of field counts did not appear to be strongly related to pedestrian volumes. Situations were identified when field counts and video counts are best applied. While field counts have a higher degree of error than manual video counts, field counts allow observers to easily collect additional pedestrian data such as pedestrian characteristics and behavior. It is suggested that manual video data reduction is best used in situations when the accuracy of the count is a primary goal.

Citation:

Mondheim, R., Methodological Aspects of Surveying the Volume, Structure, Activities and Perceptions of City Centre Visitors. In GeoJournal 45, Kluwer Academic Publishers, 1998, pp. 273-287.

Pedestrian Consideration	Addressed?	Bicycle Consideration
Permanent Counts		Permanent Counts
Special Counts		Special Counts
Factoring Methods	X	Factoring Methods
Variance of Data		Variance of Data
Minimum Counts Needed	X	Minimum Counts Needed
Reporting Information/Customer Needs	X	Reporting Information/Customer Needs
Utilizing Different Sources		Utilizing Different Sources
Formats for Counts	X	Formats for Counts
Reasons for Data Collection	X	Reasons for Data Collection

Summary:

While the focus of this research is on using pedestrian data and opinions to determine the potential of retail locations in downtown areas, as opposed to pedestrian transportation, it does devote significant attention to pedestrian counts and associated temporal variations (time of day, day of week, and annually). It is noted that these variations are further complicated by different patterns based on the study site's physical location and the function of the adjacent street (for example, a street in an office district with prevalent pedestrian commuting versus a shopping district street with a more uniform time of day pattern). The research includes a discussion of a “representative weekday” that can be used to extrapolate daily counts to weekday counts, with a specific comparison of Tuesday counts and Saturday counts at sites worldwide. The second half of the paper includes a discussion of pedestrian interviews (i.e., surveys). Again, while the context is asking pedestrians about their opinions regarding city centers and shopping districts, there is some useful discussion regarding weighting of surveys both temporally and spatially.

Citation:

Schneider, R.J., L.S. Arnold and D.R. Ragland. Methodology for Counting Pedestrians at Intersections: Use of Automated Counters to Extrapolate Weekly Volumes from Short Manual Counts. In Transportation Research Record: Journal of the Transportation Research Board, No. 2140, Transportation Research Board of the National Academies, Washington, D.C., 2009, pp. 1-12.

Pedestrian Consideration	Addressed?	Bicycle Consideration
Portable Counts	X	Portable Counts
Permanent Counts	X	Permanent Counts
Special Counts		Special Counts
Factoring Methods	X	Factoring Methods
Variance of Data	X	Variance of Data
Minimum Counts Needed		Minimum Counts Needed
Reporting Information/Customer Needs	X	Reporting Information/Customer Needs
Utilizing Different Sources	X	Utilizing Different Sources
Formats for Counts	X	Formats for Counts
Reasons for Data Collection	X	Reasons for Data Collection

Summary:

This presents a method for using manual and automatic pedestrian counts to estimate weekly intersection pedestrian volumes. Included is an outline of the specific methods for intersection selection (to provide a variety of intersection types and environmental conditions), manual counts, automatic counts, and how to manage variances in the data associated with temporal, spatial and weather factors.

Citation:

Schneider, R. J, L.S. Arnold and D.R. Ragland. Pilot Model for Estimating Pedestrian Intersection Crossing Volumes. In Transportation Research Record: Journal of the Transportation Research Board, No. 2140, Transportation Research Board of the National Academies, Washington, D.C., 2009, pp. 13-26.

Pedestrian Consideration	Addressed?	Bicycle Consideration
Portable Counts	X	Portable Counts
Permanent Counts	X	Permanent Counts
Special Counts		Special Counts
Factoring Methods	X	Factoring Methods
Variance of Data	X	Variance of Data
Minimum Counts Needed		Minimum Counts Needed
Reporting Information/Customer Needs	X	Reporting Information/Customer Needs
Utilizing Different Sources	X	Utilizing Different Sources
Formats for Counts	X	Formats for Counts
Reasons for Data Collection	X	Reasons for Data Collection

Summary:

This presents a pilot model for using manual and automatic pedestrian counts to estimate weekly intersection pedestrian volumes. Additionally, the authors outline the specific methods for intersection selection (to provide a variety of intersection types and environmental conditions), manual counts, automatic counts, and how to manage variances in the data associated with temporal, spatial and weather factors.

BICYCLE COUNTS

Information on bicycle data processing and storage is largely found through case studies. The information for data processing and storage in these case studies generally falls into categories of factoring methods, count variance, or sampling

The international literature search did not reveal significant information on bicycle data recording formats or file structures. While researchers and practitioners whose work was reviewed for this literature search must have developed recording formats and file structures for their analyses, these were not discussed in any detail in the papers reviewed. Presumably, further exploration and follow-up correspondence with the authors would allow for the collection of format and structure information.

Factoring Methods

Factoring methods help reduce error and variances in data. Additionally, factoring methods are used to extrapolate findings from sites to similar locations, larger geographies or generalize data for modeling purposes. Factoring methods typically use a baseline data set, a sample data set of field data and statistical methods to process data. Manual counts, video processing by a person, or seasonal counts are typically used to establish a baseline data set or factoring measure.

As with the pedestrian mode, significant portions of the international literature are devoted to methods that can be used to factor counts over longer periods of time. Two frequently discussed factoring methods involve day-of-week and seasonal variations.

Studies in Australia and Norway indicated significantly higher use of bicycle facilities on weekdays, while the Swiss national cycle network sees higher usage on weekends. The Swiss research additionally observes that ideal weather conditions, defined as a temperature of approximately 25° C (approximately 77° F) with no rainfall, lead to increased bicycling activity (Richardson, 2006). Different adjustment factors have been developed to account for these temporal variations based on local conditions, but no international standard has been cited.

Count Variance

All bicycle data collection has some degree of error or bias. Count variance can be attributed to several factors including equipment technology, environmental factors, site conditions, seasonal, weekly or time of day factors, and human error.

Sampling

Sampling for bicycle data collection can be applied to site selection, day and time, or participant selection for studies that track individuals. Methods for site selection range from the use of local knowledge of facilities, facility characteristic, facility volumes, geography or spatial characteristics, such as urban or rural. Selection criteria range from the use of site conditions criteria to statistical methods. Day and time selections are often made to capture peak periods for pedestrian volumes and to capture seasonal or weekday variations.

London, United Kingdom's Department for Transport has identified the number of counts required to accurately detect a change in various bicycle flows (Department for Transport, 1999, pg. 66).

Citation:

Monitoring Local Cycle Use. Department of Transport, London, Great Britain, 1999.

Pedestrian Consideration	Bicycle Consideration	Addressed?
Portable Counts	Portable Counts	X
Permanent Counts	Permanent Counts	X
Special Counts	Special Counts
Factoring Methods	Factoring Methods	X
Variance of Data	Variance of Data	X
Minimum Counts Needed	Minimum Counts Needed	X
Reporting Information/Customer Needs	Reporting Information/Customer Needs
Utilizing Different Sources	Utilizing Different Sources	X
Formats for Counts	Formats for Counts
Reasons for Data Collection	Reasons for Data Collection

Summary:

This traffic advisory leaflet discusses methods of counting bicyclists. It also recommends specific periods to count, time of year, and the number of days. The number of counts required to accurately detect a change in flows is also discussed.

DATA USE

Data may be collected for trends analysis, projects or to fulfill a regulatory mandate which may or may not support funding for infrastructure construction, improvements or maintenance. This section contains most of the case study documentation encountered in this review.

Bicycle and pedestrian data can be applied to a variety of situations and a variety of scales. From local communities to national agencies, bicycle and pedestrian data is used to document facility performance, assess need for bicycle and pedestrian facilities, improve safety, inform design and construction projects and inform policy and funding decisions. However, there is an expressed demand and need for a higher quantity and quality of data, as well as better standards and methods, to inform the list of applications above.

Amongst the international literature, much more focus is placed upon process and techniques for collecting data as opposed to reporting and using data. Limited examples of data use are provided below in this section.

Citation:

Jones, M.G., S. Ryan, J. Donlon, L. Ledbetter, D.R. Ragland and L. Arnold. Seamless Travel: Measuring Bicycle and Pedestrian Activity in San Diego County and its Relationship to Land Use, Transportation, Safety, and Facility Type. Publication UCB-ITS-PRR-2010-12. California PATH, ITS-Berkeley, California Partners for Advanced Transportation Technology, Institute of Transportation Studies, University of California Berkeley, 2010.

Pedestrian Consideration	Addressed?	Bicycle Consideration	Addressed?
Portable Counts	X	Portable Counts	X
Permanent Counts	X	Permanent Counts
Special Counts		Special Counts	X
Factoring Methods		Factoring Methods
Variance of Data	X	Variance of Data	X
Minimum Counts Needed	X	Minimum Counts Needed	X
Reporting Information/Customer Needs	X	Reporting Information/Customer Needs	X
Utilizing Different Sources	X	Utilizing Different Sources	X
Formats for Counts	X	Formats for Counts	X
Reasons for Data Collection	X	Reasons for Data Collection	X

Summary:

The report summarizes an extensive research project focused on bicycle and pedestrian data collection and analysis in San Diego, CA. The report summarizes current count and survey methods, describes in detail the data collection process for the project, documents the count and survey results, and presents the results of several bicycle and pedestrian models using the data collected. The review of count and survey methodology describes manual count methods, automatic counting technologies, bicycle and pedestrian survey methods, and reviews several bicycle and pedestrian models that have been developed to use the data. The data collection section describes how manual and automatic counts for bicyclists and pedestrians can be structured to collect year round data, as well as how a survey can be used to supplement the count data. The results section presents the findings from the data collection and includes information on bicycle and walking trip purpose, frequency of bicycling and walking, destination information, reasons preventing people from bicycling or walking more often, preferred bicycle and pedestrian facilities, socio-economic data, volume data, capacity data, temporal patterns, mode split and level of service information. The final section on bicycle and pedestrian models tests several predictive models. The modeling results are compared with the data collected to identify if the models are effective in estimating and forecasting bicycling and walking. The report concludes with several additional resources including a link to the count database, a training manual for bicycle and pedestrian counting, instructions for sending future data, details about the bicycle and pedestrian model, a summary of comparison surveys, and a summary of the background data used for analysis.

PEDESTRIAN STUDIES

The primary focus of the pedestrian studies reviewed focused on how counting data can be used to identify trends, analyze facility performance or document facility capacity. Information on how data is used to inform modeling and how mixing data can be used to extrapolate information was also found.

Several data gaps were highlighted throughout the literature related to data use including:

How to blend survey data with count data to extrapolate trends, capacity or design needs
Information on how bicycle and pedestrian data can be used to inform policy and funding decisions
Information on safety data collection and analysis
Guidelines for how count data can be used to improve facility design and safety improvements
Before and after studies related to new facilities or facility improvements for pedestrians

Recurring

Several resources examined how data collection programs can be used to analyze walking trends for a particular facility or geography. Case studies were not found that describe how pedestrian count data was used to improve safety or how data has been used to address funding or legislative mandates.

Trends

Two case studies were found that illustrate how pedestrian data can be used to analyze trends. Data typically includes a mix of manual counts, automatic counts and surveys to develop information about pedestrian trends.

The City of London, England began its pedestrian counting program simply to better understand the extent and nature of walking activity (Little, 2008, pg. 52). The initial counts were performed to get a baseline measure, presumably with the intent to monitor trends over time.

Safety

No case studies were found on pedestrian safety data collection as part of the literature search for this report.

Funding/Legislative Mandates

No U.S. case studies were found on pedestrian funding/legislative mandate data collection as part of the literature search for this report.

International transportation agencies, including state and municipal agencies in Australia and New Zealand, were found to regularly conduct bicycle counts to monitor system or facility performance and the benefits of investments in those bicycle facilities. In this sense, the count programs are used to justify government expenditure and perhaps gain funding for additional facilities. Switzerland has gone one step further by using additional surveys to determine ancillary investments made by users of the national cycle network (Richardson, 2006, pg. 54).

Project Specific

No case studies were found on project specific data collection to address facility construction or traffic warrants.

Research

Case studies were found on how pedestrian data can be applied to analyze capacity. No case studies were found that address quality of service and no special needs case studies were found either.

Capacity

Pedestrian data has been used to assess existing facility performance as well as extrapolate capacity need for facilities. Pedestrian data that addresses capacity has also been used to develop transportation models for local communities and regions.

See resources by Jones (2010, pg. 68), Schneider (2009, “Methodology”, pg. 62), Schneider (2009, “Pilot”, pg. 63), Schneider (2005, pg. 44), Cottrell (2003, pg. 46) and Hudson (2010, pg. 74) as well as Diogenes (2007, pg. 60).

Quality of Service

No case studies were found on pedestrian quality of service data collection, as it relates to monitoring counts or activity levels as part of the literature search for this report.

Special Needs

No case studies were found on pedestrian special needs data collection as part of the literature search for this report.

BICYCLE STUDIES

The primary focus of the bicycle studies reviewed focused on how counting data can be used to identify trends, analyze facility performance or document facility capacity. Information on how data is used to inform modeling and how mixing data can be used to extrapolate information was also found. Additionally, several studies address the collection and analysis of bicycle and pedestrian data.

Several data gaps were highlighted throughout the literature related to data use including:

How to blend survey data with count data to extrapolate trends, capacity or design needs
Information on how bicycle and pedestrian data can be used to inform policy and funding decisions
Information on safety data collection and analysis
Guidelines for how count data can be used to improve facility design and safety improvements
Before and after studies related to new facilities or facility improvements for bicyclists

No case studies were found that documented how bicycle counting data has been used to address safety or funding/legislative mandates. However, the need for this information is a common theme throughout the literature reviewed, especially as it relates to funding and legislative mandates.

Recurring

Several resources examined how data collection programs can be used to analyze biking trends for a particular facility or geography. Case studies were not found that describe how bicycle count data was used to improve safety or how data has been used to address funding or legislative mandates.

Trends

Several case studies were found that illustrate how bicycle data can be used to analyze trends. Data typically includes a mix of manual counts, automatic counts and surveys to develop information about bicycle trends.

See resources by Jones (2010, pg. 68), Jones (2006, pg. 48), Dill (2008, pg. 30), Harvey (2008, pg. 31), Charlton (2011, pg. 29), Lovejoy (2011, pg. 80) as well as the resources by Berkow (2009, pg. 72), Casello (2011, pg. 73) and Hudson (2010, pg. 74).

Citation:

Berkow, M. Using Bicycle Count Data to Measure Use of Existing Bicycle Facilities in Portland, Oregon. Presented at the 88th Annual Meeting of the Transportation Research Board, Washington, D.C., 2009.

Pedestrian Consideration	Bicycle Consideration	Addressed?
Portable Counts	Portable Counts
Permanent Counts	Permanent Counts	X
Special Counts	Special Counts	X
Factoring Methods	Factoring Methods	X
Variance of Data	Variance of Data	X
Minimum Counts Needed	Minimum Counts Needed	X
Reporting Information/Customer Needs	Reporting Information/Customer Needs	X
Utilizing Different Sources	Utilizing Different Sources	X
Formats for Counts	Formats for Counts	X
Reasons for Data Collection	Reasons for Data Collection	X

Summary:

This is an analysis of the City of Portland's methods for bicycle count data collection and how the data is managed. The information was used to identify ways to organize and analyze data. It identified how bicycle count data can be used to assess the use of bicycle facilities, inform policies and justify investment in bicycle infrastructure. Ways that Portland can improve its data collection program and minimum requirements that are needed for municipal bicycle count programs to be effective were presented.

Citation:

Casello, J., A.O. Nour, K.C. Rewa and J. Hill. Analysis of Stated-Preference and GPS Data for Bicycle Travel Forecasting. Presented at the 90^th Annual Meeting of the Transportation Research Board, Washington, D.C., 2011.

Pedestrian Consideration	Bicycle Consideration	Addressed?
Portable Counts	Portable Counts	X
Permanent Counts	Permanent Counts
Special Counts	Special Counts	X
Factoring Methods	Factoring Methods	X
Variance of Data	Variance of Data	X
Minimum Counts Needed	Minimum Counts Needed	X
Reporting Information/Customer Needs	Reporting Information/Customer Needs	X
Utilizing Different Sources	Utilizing Different Sources	X
Formats for Counts	Formats for Counts	X
Reasons for Data Collection	Reasons for Data Collection	X

Summary:

Two data collection methods were used, an on-line survey and GPS units to collect bicycling trip data, to produce trip generation and attraction rates for cycling as a function of land use. The mixed-data analysis method was used to inform models for bicycle trip distribution, mode and path choice, as well as prioritization of infrastructure investments.

Citation:

Hudson, J. G., T.T. Qu and S. Turner. Forecasting Bicycle and Pedestrian Usage and Research Data Collection Equipment. Publication TTI No. P2009330. Texas Transportation Institute, Capital Area Metropolitan Planning Organization and the Federal Highway Administration, 2010.

Pedestrian Consideration	Addressed?	Bicycle Consideration	Addressed?
Portable Counts	X	Portable Counts	X
Permanent Counts	X	Permanent Counts	X
Special Counts	X	Special Counts	X
Factoring Methods	X	Factoring Methods	X
Variance of Data	X	Variance of Data	X
Minimum Counts Needed	X	Minimum Counts Needed	X
Reporting Information/Customer Needs	X	Reporting Information/Customer Needs	X
Utilizing Different Sources	X	Utilizing Different Sources	X
Formats for Counts	X	Formats for Counts	X
Reasons for Data Collection	X	Reasons for Data Collection	X

Summary:

This is an analysis of bicycle and pedestrian monitoring programs around the United States, testing equipment used to collect data for bicycles and pedestrians travel, identification of best practices for data collection and storage, and identification of ways bicycle and pedestrian data can be used to inform regional transportation models and project selection criteria. Included is a discussion on how bicycle and pedestrian monitoring programs can help answer questions related to whether programs are actually increasing the number of people who bike and walk.

Project Specific

A case study was found that documents the ridership volumes before and after a facility as constructed.

An Australian study was found which used bicycle data to build models to forecast use as a function of climate variables.

Facility Construction

A study in St. Petersburg, Florida documented the before and after bicycle volumes to see whether the facility supported an increase in riders.

Citation:

Hunter, W. W., R. Srinivasan and C. Martell. Change in Amount of Bicycling Associated with Installation of Bike Lanes in St. Petersburg, Florida. Presented at the 89th Annual Meeting of the Transportation Research Board, Washington, D.C., 2010.

Pedestrian Consideration	Bicycle Consideration	Addressed?
Portable Counts	Portable Counts	X
Permanent Counts	Permanent Counts
Special Counts	Special Counts	X
Factoring Methods	Factoring Methods	X
Variance of Data	Variance of Data	X
Minimum Counts Needed	Minimum Counts Needed	X
Reporting Information/Customer Needs	Reporting Information/Customer Needs	X
Utilizing Different Sources	Utilizing Different Sources
Formats for Counts	Formats for Counts	X
Reasons for Data Collection	Reasons for Data Collection	X

Summary:

This is a case study of how to use bicycle counts to analyze whether the installation of a bike lane influenced bicycle ridership. A method for data collection is presented that includes seasonal data collection, equipment, and equipment placement.

Citation:

Ahmed, F. and C. Jacob. Impact of weather on commuter cyclist behavior and implications for climate change adaptation. Presented at 33rd Australian Transportation Research Forum, Canberra, Australia, 2010.

Pedestrian Consideration	Bicycle Consideration	Addressed?
Permanent Counts	Permanent Counts
Special Counts	Special Counts
Factoring Methods	Factoring Methods
Variance of Data	Variance of Data
Minimum Counts Needed	Minimum Counts Needed
Reporting Information/Customer Needs	Reporting Information/Customer Needs
Utilizing Different Sources	Utilizing Different Sources
Formats for Counts	Formats for Counts
Reasons for Data Collection	Reasons for Data Collection	X

Summary:

This paper did not discuss methods for obtaining bicycle counts. Rather it discussed research which used the counts to develop models to explain variations in bicycle commuting as a function of weather related variables: wind speed, daily precipitation, apparent temperature, hours of sunshine, and humidity.

The modeling effort made use of Vic Roads bicycle count stations. Subsequent research obtained from the Vic Roads website is provided in a separate summary.

Research

Case studies were found on how bicycle data can be applied to analyze capacity as well as unique bicycle data applications.

Capacity

Bicycle data has been used to assess existing facility performance as well as extrapolate capacity need for facilities. Pedestrian data that addresses capacity has also been used to develop transportation models for local communities and regions.

See resources by Jones (2010, pg. 68) and Hudson (2010, pg. 74).

Special Needs

In addition to on-street and path facilities bicycle counting data, two resources presented unusual methods to collect and analyze bicycle data. One resource analyzed the use of bicycle parking facilities to extrapolate information on bicycle parking behavior, effectiveness of bicycle parking facilities and trip and origin destinations. The other study used mixed data collection methods to extrapolate bicycle use and behavior data.

Citation:

Constant, A., A. Messiah, L. Felonneau and E. Legarde. Investigating Risk Compensation Theory in Cyclists: Results from Intelligent Video System. Presented at International Conference on Safety and Mobility of Vulnerable Road Users: Pedestrians, Motorcyclists, and Bicyclists, Jerusalem, Israel, 2010.

Pedestrian Consideration	Bicycle Consideration	Addressed?
Portable Counts	Portable Counts	X
Permanent Counts	Permanent Counts
Special Counts	Special Counts	X
Factoring Methods	Factoring Methods
Variance of Data	Variance of Data	X
Minimum Counts Needed	Minimum Counts Needed
Reporting Information/Customer Needs	Reporting Information/Customer Needs	X
Utilizing Different Sources	Utilizing Different Sources
Formats for Counts	Formats for Counts	X
Reasons for Data Collection	Reasons for Data Collection	X

Summary:

An observational study was undertaken in Bordeaux, France to determine whether helmet use among bicyclists leads to riskier bicycling behaviors (in accordance with “risk compensation theory”). An Intelligent Video Analysis System (IVAS), consisting of an Internet Protocol camera (used to detect and measure the speed of bicyclists) and a synchronized camera that takes pictures to record other characteristics including helmet use were used in the study. The study compares risky behaviors between men and women and helmeted cyclists and non-helmeted cyclists. “Risky” behaviors (including higher speed travel, on-road riding, and running of red lights) were more commonly observed among helmeted males than non-helmeted males, but no significant differences were discovered among the female population. While the IVAS may have additional applications for bicycle and pedestrian counting, it is not described in any significant detail in the research summary.

Citation:

Lovejoy, K. and S.L. Handy. Mixed Methods of Bike Counting for Better Cycling Statistics: The Example of Bicycle Use, Abandonment, and Theft on UC Davis Campus. Presented at the 90th Annual Meeting of the Transportation Research Board 90th Annual Meeting, Washington, D.C., 2011.

Pedestrian Consideration	Bicycle Consideration	Addressed?
Portable Counts	Portable Counts	X
Permanent Counts	Permanent Counts
Special Counts	Special Counts	X
Factoring Methods	Factoring Methods	X
Variance of Data	Variance of Data	X
Minimum Counts Needed	Minimum Counts Needed	X
Reporting Information/Customer Needs	Reporting Information/Customer Needs	X
Utilizing Different Sources	Utilizing Different Sources	X
Formats for Counts	Formats for Counts	X
Reasons for Data Collection	Reasons for Data Collection	X

Summary:

This documents use of three different methods of bicycle data collection to project bicycle travel behavior and total volumes of bicycle ridership at the University of California-Davis. The data sources used include police records on reported bike thefts, a bike rack count, and a travel survey. The sample results from the surveys, reports and counts were used to make projections about bicycle use for the entire campus population. The combination of measurement methods provided a more robust picture of bicycle volumes than any one method alone could.

Citation:

Moskovitz, D. A. and N. Wheeler. Bicycle Parking Analysis Using Time-Series Photography. Presented at the 90th Annual Meeting of the Transportation Research Board, Washington, D.C., 2011.

Pedestrian Consideration	Bicycle Consideration	Addressed?
Portable Counts	Portable Counts
Permanent Counts	Permanent Counts
Special Counts	Special Counts	X
Factoring Methods	Factoring Methods	X
Variance of Data	Variance of Data	X
Minimum Counts Needed	Minimum Counts Needed
Reporting Information/Customer Needs	Reporting Information/Customer Needs	X
Utilizing Different Sources	Utilizing Different Sources
Formats for Counts	Formats for Counts	X
Reasons for Data Collection	Reasons for Data Collection	X

Summary:

This presents a method for collecting and analyzing bicycle parking data. Digital photography was used to capture bicycle parking data over a period of time that includes arrival and departure times, parking durations and turnover rates. The data can be used to answer questions related to bicycle parking behavior, effectiveness of bicycle parking facilities, and trip and origin destinations.

Citation:

Phung, J. and G. Rose. Temporal Variations in Usage of Melbourne's Bike Paths. Presented at 30th Australasian Transport Research Forum, Melbourne, Australia, 2007.

Pedestrian Consideration	Bicycle Consideration	Addressed?
Permanent Counts	Permanent Counts
Special Counts	Special Counts
Factoring Methods	Factoring Methods
Variance of Data	Variance of Data
Minimum Counts Needed	Minimum Counts Needed
Reporting Information/Customer Needs	Reporting Information/Customer Needs
Utilizing Different Sources	Utilizing Different Sources
Formats for Counts	Formats for Counts
Reasons for Data Collection	Reasons for Data Collection	X

Summary:

This paper did not discuss methods for obtaining bicycle counts. Rather it discussed research which used the counts to develop models to explain variations in bicycle commuting and recreational use as a function of temporal variables: time of day, day of week, month of year. Based upon their analysis the researchers recommend hour to day, day to week, and month to year adjustment factors.

The modeling effort made use of Vic Roads bicycle count station. Subsequent research obtained from the Vic Roads website is described in a separate citation.

Citation:

Ellison, R, and S. Greaves. Travel Time Competitiveness of cycling in Sydney. Presented at 33rd Australasian Transport Research Forum, Canberra, Australia, 2010.

Pedestrian Consideration	Bicycle Consideration	Addressed?
Permanent Counts	Permanent Counts
Special Counts	Special Counts	X
Factoring Methods	Factoring Methods
Variance of Data	Variance of Data
Minimum Counts Needed	Minimum Counts Needed
Reporting Information/Customer Needs	Reporting Information/Customer Needs
Utilizing Different Sources	Utilizing Different Sources
Formats for Counts	Formats for Counts
Reasons for Data Collection	Reasons for Data Collection	X

Summary:

This paper did not address counting of bicyclists for use in determining volumes. Rather, GPS units were provided to cyclists and motorists accessing areas and specific destinations to compare travel times among cyclists and motorists.

REPORTING

Reporting is an outcome of a defined need to collect information. In assigning material to this section three questions were asked: Does this material provide information on standardizing a record format for data storage? Is the method of presentation “easily” transferrable? Is there a mandate for this information that a consistent presentation of the material would simplify comparison or aggregation of the information provided.

The standards identified as part of this literature review are based on the standards established by the National Bicycle and Pedestrian Documentation Project. The project's standards have evolved over time and have also been adapted to local needs, methods and programs.

The standards include record formats for intersection counts, screenline counts and surveys. The standards allow for facilities types to be compared and data aggregated to analyze larger geographic areas.

See resources by Jones (2010, pg. 68), Jones (2006, pg. 48) and the Metropolitan Transportation Commission (2003, pg. 50).

Reporting standards for safety, funding/legislative mandates, facility construction and traffic warrants were not found.

Amongst the international literature, much more focus is placed upon process and techniques for collecting data as opposed to reporting and using data.

VicRoads (Melbourne, Australia) maintains a website which includes data collected. Bicycle count information is reported by time of day, day of week, and month of year. Each year a narrative analysis is prepared of the bicycle count data. These analyses present hypothesis to explain some variations from the norm in annual data (e.g. a particular month was particularly rainy).

[1] Time Series photography (parking racks)

[2] Police records (thefts); counts at bike racks, On-line travel survey by invitation to stratified sample

Return to top

Page last modified on October 19, 2015