Rideshare Operations Management for Heterogeneous Vehicles to Minimize Whole CO2 Emissions

Abstract

"In recent years, it has become increasingly necessary to reduce CO2 emissions as a countermeasure to global warming. Paris Agreement sets out a global framework to limit global warming to well below 2degC, which needs an extremely huge reduction of CO2 emissions. Therefore, it is critical to deal with the use of automobiles that has a big impact on this problem. On the other hand, since it is difficult to dramatically improve individual vehicle efficiency anymore, new approaches are required to achieve the goal. Because of this background, we, DENSO, are proceeding with development to reduce CO2 emissions not only by performance improvement of individual vehicles but also by operations management or cooperative control of multiple vehicles. In this paper, a rideshare operations management technology reducing the total fuel spend is introduced, where several users going to the common destination share a ride in the same vehicle. In general, it is well known that whole CO2 emissions from multiple vehicles can be reduced if the use of vehicles is reduced with ridesharing. Meanwhile, there is room for an additional reduction of CO2 emissions because, in existing researches, only simple cost functions such as the sum of route costs are minimized without considering vehicle efficiency. Besides, since the vehicles used for rideshare are often heterogeneous with various kinds of powertrain systems, CO2 emissions can be reduced further by taking these characteristics. Therefore, a simulation model including each powertrain characteristics for rideshare operations management is proposed to calculate CO2 emission from each vehicle for each driving route precisely. Then, an optimization problem is formulated by designing a cost function described with the total amount of CO2 emissions calculated with the simulator. Since this problem is complicated, where not only a destination but also users going to the destination must be assigned to each vehicle, mixed-integer linear programming (MILP) is not an appropriate methodology to find the optimal solution. Therefore, an operations management technology has been developed to find the optimal destination and user allocation by incorporating this problem into a constraint satisfaction problem (CSP) for minimizing the total amount of CO2 emissions from multiple vehicles. The effect of this technology has been validated with simulation, showing more than 30% CO2 emissions reductions in some cases comparing with conventional rideshare technologies, where the amount of used vehicles is minimized."