The Exploratory Advanced Research Program

Vehicle Positioning, Navigation, and Timing: Leveraging Results From EAR Program-Sponsored Research

WORKSHOP SUMMARY REPORT   •   November 2012

Introduction

On November 20, 2012, at the Turner-Fairbank Highway Research Center (TFHRC) in McLean, VA, the Federal Highway Administration's (FHWA) Exploratory Advanced Research (EAR) Program and Office of Operations Research and Development (R&D) convened a workshop to share information about the results of EAR Program-sponsored research on vehicle positioning and navigation.

The workshop, titled "Vehicle Positioning, Navigation, and Timing: Leveraging Results From EAR Program-Sponsored Research," was held to identify key government, industry, and academic audiences who would be interested in the results and how the EAR Program can assist in connecting the audiences with the results. It provided an opportunity to discuss potential follow on applied areas of research in addition to addressing continued fundamental research gaps that still need to be resolved to provide dependable, precise, and commercially affordable positioning and navigation for roadways.

The audience included Government program managers and researchers involved in the research, development, deployment, or regulation of positioning and navigation for increased safety, mobility, and efficiency in transportation systems.

The workshop began with a welcome from Joe Peters, Director, Office of Operations R&D at TFHRC. Peters described the work of the Saxton Transportation Operations Laboratory at TFHRC. He stated that the Saxton Laboratory includes exploratory work in automated systems and described the three teams that comprise the Office of Operations R&D as follows:

• Transportation Enabling Technologies.
• Transportation Operations Concepts and Analysis.
• Transportation Operations Applications.

This Vehicle Positioning, Navigation, and Timing Workshop was sponsored by the Transportation Enabling Technologies Team. Peters highlighted the vital importance of position, navigation, and timing (PNT) technologies. As an example, without a global positioning system (GPS) signal providing a reliable timing reference, traffic signals would lose the ability to function in a coordinated way, resulting in widespread traffic jams.

David Kuehn, EAR Program Manager, stated that FHWA's research is mission-driven to deliver mobility to the Nation's highways. The EAR Program looks for solutions beyond transportation, exploring the fields of science and engineering for next-generation applications, but remains ever mission-focused.

Several EAR Program-funded projects are related to positioning and navigation. The results of two of these projects were presented at the workshop and are detailed in this report. Although these projects are highway-focused, this workshop serves as an opportunity to apply these results to other transportation modes.

An overview of the two projects was then presented. Jim Arnold introduced the University of California at Riverside's research on new approaches in vehicle positioning. He stressed that one of the most importation aspects in research is a better definition of requirements.

David Gibson introduced Auburn University's research on vehicle positioning in GPS-degraded environments. He emphasized the importance for having calibrated test tracks in test environments.

Gary Pruitt of ARINC and Scott Andrews of Cogenia Partners then presented a talk on "Position Technology and Requirements for Connected Vehicles." An assessment on accuracy and positioning errors was presented. The focus of the talk was mainly on positioning errors and methods to improve accuracy.

Pruitt and Andrews attempted to define a process for systematically deriving requirements for positioning requirements for a given application. To accomplish this, there needs to be a better understanding of positioning errors and methods for improving position estimates. There are two main types of errors—random and offset.

• Random position errors cause position estimates to change from measurement to measurement. These errors are specified in terms of a statistical representation.

• Offset errors will affect all receivers of the same type in the same general manner. These errors are either systematic or ionospheric. They can be minimized by using differential GPS.

Other errors were also discussed, including motion errors. From moving platforms, a person only gets one sample at a given point instead of at a distribution of points. Latency errors are small time errors between platforms. These can be fixed with synchronized clocks. There may also be a processing lag caused by the use of algorithms to help smooth variations in position measurements. Map errors—where landmark positions are incorrect—were also discussed.

The talk then focused on decision errors, in which measurement and reality are compared. Correct decisions are made when measurement and reality are the same, whereas decision errors occur when measurement and reality are not the same. These error conditions cause false positives and false alarms. A false positive (i.e., when the measurement indicates true when in fact the reality is false) can lead to a missed detection or a dangerous situation. A false negative (i.e., when the measurement indicates false when in fact the reality is true) can lead to a nuisance or false alarm.

Next presented was a discussion of error tolerance and tolerable errors. In practical systems, some level of a tolerable error needs to be defined. These tolerances are different for false positives and false negatives, with a much larger tolerance for false positives.

The presentation then focused on hazard analysis, in which severity, exposure, and controllability determination were discussed. Severity was defined as a range that spanned from no injuries to life-threatening injuries, the term exposure varied from an incredibly low to a high chance of occurrence, and controllability indicated a range that spanned from generally controllable to difficult to control. Severity, exposure, and controllability determinations could be measured by a probability of occurrence.

The talk concluded with a summary of requirements and a summary of positioning technology. The presenters recommend the development of a decision–error method for determining appropriate positioning requirements. They also identified additional requirements based on dynamic factors and suggested time syncing to determine relative positioning. The summary of the presenters’ review of positioning technologies identified GPS as the most promising (other technologies are less accurate and require significant infrastructure). They stressed that there is an emerging interest by GPS makers to explore low-cost, high-accuracy vehicle solutions.

Several questions were asked in response to the presentation. One question pertained to the scale of production or potential market for purchasing positioning technology. The response involved a discussion on receivers used for survey instruments. There is marginal additional cost for the hardware, but companies may charge more for additional performance as a way of recouping development costs.  For example, some receivers are sold with full capabilities and users purchase licenses to activate the desired features. Further discussion pertained to errors and latency and how to develop budgets for cumulative errors.