Project: #4 Optimizing Snow Plowing Operations in Urban Road Networks Progress Report - Reporting Period Ending: March 31, 2018 Principal Investigator: Stephen Smith Status: Overdue Project Start Date: Jan. 1, 2017 End Date: April 30, 2018 Research Type: Applied Grant Type: Research Grant Program: Private Funding Grant Cycle: 2017 Traffic21 Progress Report (Last Updated: March 31, 2018, 6:22 a.m.) % Project Completed to Date: 95 % Grant Award Expended: 95 % Match Expended & Document: 25 USDOT Requirements Accomplishments During the last reporting procedure we have achieved two major accomplishments: (1) In-Vehicle app - The in-vehicle app for providing turn-by-turn instructions was re-engineered to rely on Map Box as the underlying mapping tool to overcome limitations in the interface to Google maps. Extensions were also made to implement a "skip" button, for purposes of deviating from the planned route in the event that it is impassible (e.g., due to an abandoned vehicle). When invoked the system will reroute the vehicle back onto the planned route as soon as possible. In January 2018, the app was successfully pilot tested with a City of Pittsburgh vehicle driver from Division 3 - driving a current City route in division 3. (2) Dynamic route planner - Further improvements were made to the dynamic route planning system that we have developed, and infrastructure was developed to import City of Pittsburgh route and vehicle data for use in generating routes. Using city data used to generate the City's current snow plow routes for the Green field neighborhood, an initial performance analysis of the CMU route planner was performed. With both sets of routes using 3 10-ton vehicles, the CMU planner was able to generate routes that finish 13 minutes sooner that the current City routes (1h/27m versus 1h/40m), with 1 fewer u-turn than the City routes. Primary routes are cleared 12 minutes earlier; secondaries about 11 minutes earlier. Impacts Both the successful field test of the in-vehicle app and the comparative improvement shown by the CMU route planner over the City's current snow plow routes demonstrate the potential for significant improvement of City operations. Whereas our initial comparison of generated routes already shows approximately a 13% reduction in plowing time, we expect to be able to significantly improve over these initial results as we further refine the CMU planner's heuristics. Another important point to note is that the CMU planner can operate in an "incremental" mode, which will enable the ability to do real-time replanning of routes if the situation requires it (e.g., if a vehicle breaks down, or the storm dramatically intensifies). Such a dynamic replanning capability is not possible today. It is our intention to transition this technology to a commercial enterprise and bid on the City of Pittsburgh's current outstanding RFP for route optimization capability. Other We intend to commercialize the route planning and in-vehicle app technology we have developed, and are bidding on the City of Pittsburgh's current RFP for Fleet Telematics and Route Optimization Serves (RFP 180000193) as the means to make this happen. Outcomes New Partners None Issues None