Cell Phone Data and Travel Behavior Research: Symposium Summary Report

The Exploratory Advanced Research Program

Data Coverage, Resolution and Characteristics

In this session, industry representatives provided a brief overview of their companies and the types of data they collect and are able to provide. Symposium participants then discussed challenges relating to privacy and modes.

Presentations

Bill King, AirSage
Bill King of AirSage explained that the company is a data provider with contracts in place with two major cell phone carriers. King highlighted how AirSage is able to pinpoint a user’s location by triangulating between multiple towers, as shown in figure 1. He noted users are anonymous, but AirSage is able to link to demographics at the census tract level. AirSage is also able to make inferences about users’ home and work locations within a dataset of extrapolated person trips. King stated that the data can be licensed to end users for a specific purpose.

Figure 1. AirSage pinpoints a user's location.

Daniel Rolf and Keith Hangland, HERE
Daniel Rolf and Keith Hangland of HERE informed symposium participants that, although originally a mapping company, HERE now uses probe data collected from different devices by using applications or cell phones. Rolf and Hangland noted that HERE is not only collecting and selling probe data but is also using it. They explained that HERE data power freight and trucking industries and are also beginning to be used for government planning activities. Rolf and Hangland told participants that data can also be used to analyze traffic movement and efficiency to solve engineering and planning issues.

Rick Schuman, INRIX
Rick Schuman of INRIX stated that the primary business for INRIX is real-time congestion or traffic movement data based on 25–30 percent of travelers, as shown in figure 2. Schuman noted that this can be obtained in real time or through a record or catalog. He explained that INRIX provides processed data, which is more usable for most buyers than raw data. Schuman told symposium participants that there is also an important distinction between metro-level data (generally associated with a project) and national-level data (generally associated with a product).

Figure 2. Cellular analytics illustrate where a region's visitors originated and highlights areas of demand.

Nick Cohn, TomTom
Nick Cohn of TomTom informed symposium participants that one feature of TomTom’s data is origin–destination (OD). Cohn noted that, like other providers, available data are a sample of the total population. He explained that sample sizes and local data availability vary greatly, making it difficult to standardize a national product. Cohn highlighted that TomTom gathers Global Positioning System (GPS) data from vehicle movements used for both speed and OD. He stated that TomTom uses OD data, along with mapping and shape files, and makes data available through licenses. TomTom also creates customized datasets for specific customers and their needs.

Major Themes Discussed

The symposium participants discussed several major themes during this session, as follows:

• Connecting data and behavior. Although land-use data are helpful and available at a city level, in terms of making connections between the location data and user behavior, there is currently no such nationwide map.
• Identifying needs. The financial and advertising industries already use OD analytics but have different needs in terms of precision and bias. For the transportation industry, it should be established what margin of error is acceptable in terms of deriving behavior.
• Confirming data usage. Licensing differs for educational facilities, researchers, government, and commercial entities who want to use available data.
• Making data available. There are privacy issues in terms of which aspects of data are available.
• Determining mode. The ability to determine the mode (e.g., car, bus, cycling, and walking) varies depending on location and context as well as the acceptable margin of error.

Key Takeaways

The symposium participants noted several key takeaways during the discussion, as follows:

• Data are available from providers through specific licenses and agreements.
• OD trips can be determined; however, the reason for a trip and mode are less certain and have wider margins for error.
• Privacy issues weigh upon data availability to some degree.