Hunting the Energy Optimal Fleet Composition 

Abstract
The fleet of heavy-duty commercial trucks includes a variety of vehicles/powertrains and is responsible for a diverse mission profiles. Even when considering only conventional Class 8 trucks, variations in powertrain components such as engine power rating, transmission type, axle ratios, tires, and vehicle customizations such as chassis type, cab type, and freight capacity, lead to many different vehicle configurations. The set of all possible configurations is referred to as the design space and a configuration with a specified set of component sizes is referred to as a candidate. In this process, users, dealers, and engineers interact with each other to come up with a vehicle configuration (candidate) that will meet the performance and cost expectations of the user fleet. However, when the goal is energy consumption reduction, it is very important to assign the right powertrain to the right mission profile. In this work, we demonstrate that the ability to predict a truck’s energy consumption beforehand can enable an optimal match between the vehicle/powertrain and the mission profile. Thus we can modulate fleet operations to better utilize energy resources and reduce the fuel consumption of the whole fleet. We are using a mix of physics and data-driven/ML models for the framework and are utilizing the historical data. The feasible candidates are evaluated and shortlisted based on their energy consumption for the mission profile. The energy footprint of the given candidate (vehicle configuration) is predicted, it is assigned to the respective mission profile for energy-efficient operations. It is expected that the energy optimal fleet composition can save upto 6% energy resoucres.