Abstract
This project proposes to configure a local signalized intersection to demonstrate and validate a low-cost approach to deployment of PedPal, a smartphone app designed to support safe intersection crossing app for persons with disabilities. Recent work with PedPal has designed and prototyped solutions to overcome practical limitations and obstacles to widespread deployment, including (1) the use of prepositioned Ultra-wide band (UWB) beacons at the intersection to overcome the inherent localization accuracy of smartphones in urban canyons and allow corner detection and pedestrian tracking at a few centimeters accuracy, (2) a broadening of capabilities to include pedestrian-friendly route planning and wayfinding and support for the ‘complete trip’, and (3) specification of a PedPal system variant that eliminates its dependence on the surtrac adaptive traffic signal system and instead is capable of operating with the conventional, fixed-timing plan, traffic signal control systems that typify most urban environments], providing a much more economical basis for PedPal deployment. These collective advances have led to a recent project award to Sacramento County within DOT’s Advanced Transportation Technologies and Innovative Mobility Deployment (ATTAIN) program to produce a hardened version of the complete trip app prototyped in that operates with Sacramento County’s Centracs Edaptive traffic control system, and carry out a large-scale two-year pilot test over a 31-intersection region of the county (nominal start date: first quarter, 2025). Our goals in this proposed project are threefold: (1) to consolidate the above mentioned advances into a hardware/software system that is deployable with a conventional traffic signal system running a fixed timing plan and demonstrate its operation, (2) to provide a local site for testing and validating components of the hardware/software system that are applicable to the Sacramento deployment, and (3) to establish a basis for subsequently deploying PedPal in Pittsburgh in the future.
In more detail, the City of Pittsburgh’s Department of Mobility and Infrastructure (DOMI) has agreed to give us access to an intersection in the vicinity of the CMU campus that currently operates with a higher-end 2070 hardware controller. Use of a 2070 controller equipped intersection will enable us to utilize the standard NTCIP communication protocol and provide a version of PedPal that will operate with any NEMA-compliant controller, ensuring a measure of generality in the solution that is developed across the controllers of different manufacturers. We will augment the equipment in the cabinet at the intersection to include a separate processor to act as a surrogate for the surtrac system in the original PedPal system architecture. This surrogate process will communicate with the PedPal cloud server on one end and invoke US DOT’s V2I-Hub software on the other end to interact with the 2070 controller. Realization of this surrogate process, which we refer to as the PedPal roadside unit (P-RSU), will constitute the proposed effort’s principal software development task. To enable WIFI connectivity between the PedPal cloud server and the P-RSU, we will also install a router in the cabinet. Finally, an Estimote UWB tag (or equivalent) will be installed at a known location at each corner of the intersection. Once the intersection is fully configured and the P-RSU is operational, we will run a series of tests to verify the performance of all app capabilities. We will provide DOMI with access to the final PedPal app for subsequent adoption. Currently, there are a number of 2070s scattered around the city and this number is likely to grow over time as the Smart Spines project progresses.
Description
Timeline
Strategic Description / RD&T
Section left blank until USDOT’s new priorities and RD&T strategic goals are available in Spring 2026.
Deployment Plan
Development and validation of a version of PedPal that operates in the field with a conventional traffic signal system running a fixed timing plan will proceed according to the following project plan.
During the first quarter, work will focus on the following tasks:
• In collaboration with the City of Pittsburgh, selection of a local test intersection equipped with a NEMA-compliant 2070 hardware controller for demonstration and validating our technology results.
• Installation of DOT’s V2I-Hub software on a separate processor to be installed in the intersection cabinet, and development of a software service for managing access to V2I-Hub. The first quarter goal will be to simply establish connectivity between this software service resident in the intersection cabinet and the PedPal cloud server.
• Determination of an appropriate location at each intersection corner for placement of UWB beacons and subsequent installation of the beacons
• Finalization, together with the UWB beacon vendor, of a plan for establishing initial connectivity with PedPal when its user arrives at the intersection that bypasses the need for reliance on Bluetooth. In our prototype demonstration of UWB viability a cellular connection was exploited, and it is anticipated that the latest version of the beacons will accommodate this approach.
During the second quarter, work will focus on the following tasks:
• The software service resident in the processer at the cabinet will be extended to integrate with the V2I hub, focusing on production of a DSRC formatted MAP message for broadcast when the PedPal server indicates that the user is in range. An internal representation of the MAP message will be generated, including intersection geometry, street names, number of lanes, lane length, locations of crosswalks and locations of corner beacons. The ability for PedPal to communicate its proximity to the intersection through the cloud server and receive back a MAP message will be demonstrated.
• A solution for establishing initial connectivity between the PedPal app and a given UWB beacon will be implemented and verified, following the plan that was formulated with the UWB vendor in the first quarter.
During the third quarter, work will focus on the following tasks:
• The software service/ V2I-Hub component will be integrated with the hardware controller, using the NTCIP protocol stipulated by the NEMA standard, to enable periodic creation and broadcast of DSRC formatted Signal Phase and Timing (SPaT) messages in response to an indication from the PedPal cloud server that the PedPal user is within range of a given intersection corner. The ability for PedPal to communicate its proximity to an intersection corner through the cloud server and receive back a MAP message and subsequent SPaT messages once a second will be demonstrated.
• The ability to accurately track user crossing progress using PedPal app communication with the UWB beacons to achieve localization accuracy will first be verified, followed by the first demonstration in the field of the ability to detect user movement outside of the crosswalk and signal corrective action. This capability was demonstrated this past summer using our traffic simulator and will likely need some refinement to operate effectively in the field.
During the fourth quarter, work will focus on the following tasks:
• We will conduct systematic crossing experiments to evaluate the correctness of the PedPal app’s intended behavior through all aspects of the user’s crossing experience, including accurate detection of arrival at the corner, correct communication of crossing options, correct communication of crossing phase information, accurate accounting of user progress (both headway and directional) during crossing and accurate detection of arrival at the destination corner.
• We will next evaluate PedPal system behavior when there are multiple PedPal users in the intersection simultaneously and if necessary, revise the PedPal cloud server’s user bookkeeping method.
Expected Outcomes/Impacts
We anticipate that the proposed research will significantly lower the cost barrier to municipalities interested in providing their local disability communities with advanced technology support for safe intersection crossing and wayfinding, by eliminating the pre-requisite need to have a sophisticated (and expensive) adaptive traffic signal control infrastructure in place and instead enabling operation of such mobility aids with conventional traffic signal systems that run fixed signal timing plans. Further, by providing a safe intersection crossing app architecture that is capable of operating with any NEMA-compliant hardware controller, the proposed work will enable deployment in a broad range of urban environments. Ultimately, the proposed work will significantly enhance the mobility of persons with disabilities in urban environments.
Expected Outputs
This project will produce a field tested and validated version of the PedPal safe intersection crossing app that is capable of operating with conventional traffic signal systems that run pre-programmed signal timing plans, which are by far the most prevalent type of traffic signal control system currently in use in urban settings. The project will also enable local refinement and testing of PedPal architectural components that will be required for subsequent remote deployment of an integrated PedPal/wayfinding app in Sacramento CA (e.g., enhanced smartphone app localization via use of UWB beacons) which is being funded through a 2024 award to Sacramento County from the US DOT ATTAIN program. Finally, the variant of PedPal that is produced will be made available to the City of Pittsburgh to provide a basis for future deployment here.
We anticipate that the technology results of the proposed project will result in an invention disclose and subsequent patent application.
TRID
The attached TRID search results indicate several research efforts that have focused on smartphone-based technology support to pedestrians with disabilities (principally vision impaired individuals), including several of our PedPal related publications. Bauer and colleagues (2015) introduced a framework improved safety through pedestrian guidance and navigation based on the use of mobile devices like smartphones. Khosrow and colleagues (2017, 2018) have demonstrated smartphone-based capabilities for issuing signal crossing requests to intersection roadside units and location information to nearby equipped vehicles in a DSRC radio environment. Liao and colleagues (2012, 2013, 2018, 2020) have for several years been evolving and testing a smartphone app designed specifically for visually impaired individuals that provides signal and intersection geometry information to its users using a custom controller-specific interface and utilizes a network of Bluetooth beacons to provide traffic and navigational information.
The proposed research is distinguishable from this body of related work in several important respects. First, it is unique in its use of UWB beacon technology to track pedestrian movement, which provides far greater (3 cm) accuracy than the Bluetooth devices, coupled Kalman filtering techniques and map-matching techniques used in other work. Second, it is the first effort to attempt to provide a controller-independent architecture that is applicable to a range of conventional traffic signal system implementations. Third, when coupled together with wayfinding capabilities as is planned for the pending deployment of the technology in Sacramento County, it provides for the first time, mobility support for the ‘complete trip’. More generally, it is the first effort to focus explicitly on the goal of wide-scale deployment.
Individuals Involved
| Email |
Name |
Affiliation |
Role |
Position |
| sfs@cs.cmu.edu |
Smith, Stephen |
Carnegie Mellon University |
PI |
Faculty - Research/Systems |
Budget
Amount of UTC Funds Awarded
$50000.00
Total Project Budget (from all funding sources)
$100000.00
Documents
Match Sources
No match sources!
Partners
| Name |
Type |
| pathVu |
Deployment Partner_ Deployment Partner_ |
| City of Pittsburgh (DOMI) |
Deployment Partner Deployment Partner |
