This Small Business Innovation Research (SBIR) Phase I project will develop an affordable integrated vehicle fleet and mobile workforce management system for improved operational efficiency and environmental sustainability by better route assignments, refueling advice, and real-time vehicle connectivity for heterogeneous fleets. This system will incorporate three novel technologies to provide an end-to-end solution to the heterogeneous fleet management problem: 1) route assignments that take into account different types of drive trains and correspondingly, different fuel consumption along different types of routes; 2) refueling advice that provides insight as to where and when a driver should refuel taking into consideration forecasted fuel prices and multiple fuel types; 3) and a new system for decision support is needed for refueling and routing strategies that minimize the cost of fuel, environmental impact, and customer satisfaction.
Reducing carbon emissions has become a societal problem that touches every business sector and individual. Different fuel types and different vehicle powertrains provide small fleet owners with the opportunity to improve their overall fleet's carbon footprint, but also add a huge level of complexity, with the additional volatility of prices of multiple fuel types. As fleets make up a large percentage of cars, a reduction of CO2 by fleets can have a significant impact on a region's CO2 reduction; This solution targets not only large fleet owners, but also enables small fleet owners to affordably analyze and reduce their carbon footprint. The company plans to build revenue by 1) using a SAAS model, smart phones, and cloud computing to enable small fleet owners to have an affordable option to for green routing and minimizing fuel costs for their fleets 2) partnerships with gas price database companies, such as OPIS and Gas buddy; 3) licensing to strategic partners in direction services for adaptive routing based on powertrain and drive style.
WayLogics provides innovative logistics, scheduling, and routing solutions for commercial vehicle fleets based on advancements in analytical modeling and information and communication technology. Our value proposition integrates vehicle telematics data with cloud-based modeling to minimize fuel consumption, environmental impact, and operating cost while increasing customer service. The WayLogics team used the NSF SBIR I grant to 1) develop and evaluate vehicle routing and vehicle to route assignments based on the unique vehicle attributes, such as powertrain, vehicle age, and fuel type; 2) evaluate the use of vehicle telematics and smart phones as a platform for collecting vehicle data; and 3) explore cloud-based optimization tools to run computationally intensive models. Additionally, we developed algorithms for fuel economy estimation that can be used for any vehicle. This more accurate fuel economy estimation is used by the routing algorithm as a part of the cost function to determine better routing assignments based on actual fuel consumption along a route instead of just average miles per gallon. Our Phase I results show that these routing algorithms provide better vehicle to route assignments and require just a fraction of the cost of the traditional vehicle tracking and routing infrastructure for fleet owners. There were three main pieces of work for which we proved feasibility. The first was in determining acceptable scan rates for collecting vehicle data, which vehicle sensors are necessary to accurately predict fuel consumption, and how well mobile phones with standard data plans performed. The second was in regard to how well we could estimate the fuel economy for many different vehicle types. The third was related to improving routing and vehicle to route assignments using both vehicle information (e.g. powertrain) and accurate fuel economy. The last was in demonstrating feasibility of running the optimization algorithms on the cloud. We showed that the scan rates observed through several hundred hours of testing were sufficient to support the instantaneous fuel economy calculation along the routes. Additionally, we had near 100% successful transmission rate of the sensors from the phone to the server, using less than 1% of a normal 5GB data plan. The core phone functionality was still available for use without any reduction in functionality. We demonstrated that when we estimate the instantaneous fuel flow from the industry standard signals available through the vehicle OBD-II interface for different vehicle types, the variability is within 3 sigma of the mean. We also developed methods and algorithms for adaptation of cumulative fuel consumption models of multiple vehicles over multiple route segments. Furthermore, we demonstrated feasibility of approaches to estimate the fuel consumption variance and how this can be included the risk-adjusted optimization of fuel consumption. For routing, we demonstrated the feasibility of routing algorithms that take into consideration vehicle powertrains or alternative fuels and accurate fuel economy estimation to result in substantial improvements in solution quality (i.e., 25-35%) from the perspective of minimization of overall fleet fuel consumption. We achieved this by explicitly incorporating vehicle-specific arc costs within the heterogeneous fleet vehicle routing problem (HFVRP). Finally, we showed that cloud computing is viable for both real-time data collection and storage and for running the routing optimization. The prototype of the system has been tested using actual data from a commercial fleet company and has demonstrated over 30% improvement in the fuel consumption.
