ABSTRACT
The powertrain of passenger cars is undergoing a revolution in terms of decarbonisation, primarily via electrification. A subsequent major step change in passenger cars is expected to be the development and subsequent consumer adoption of level 5 autonomous vehicles. Such a vehicle leads to very different vehicle level requirements, which cascade down to a potentially different powertrain solution which is what the authors of this paper have investigated. In parallel to this, the ownership model is widely anticipated to evolve at a slower pace, to one of shared mobility and Mobility as a Service (MaaS).
SAE Level 5 driving autonomy has potential impacts on vehicle powertrain sizing resulting from both passenger comfort requirements and use in MaaS applications. Impacting factors include rapid accelerations and jerks while passengers are working or suffering from motion sickness, the altered vehicle usage pattern and requirements due to MaaS. The hypothesis explored is whether these factors impact the associated rated power for which the powertrain is designed. This paper uses vehicle powertrain modelling to simulate a range of drive cycles to give the cost, mass, efficiency, and performance trade-off between existing powertrains and those that may be applicable to autonomous vehicles. It considers the powertrain sizing and performance of Battery Electric Vehicles (BEV).
Mobility as a service is a driver for change as the vehicle owner/ operator requires maximum up-time from the vehicle, whereas driver owned vehicles spend a vast majority of the time parked. Therefore, lengthy charging periods of large capacity battery packs may be damaging to the MaaS business case, compared to operating a more efficient vehicle that spends less time charging. However, a downsized powertrain may take longer to complete a given journey. This paper compares journey time and efficiency using an array of bespoke distance-based drive cycles; this methodology then allows vehicle uptime and running costs to be compared over a range of anticipated and idealised demands.
Using a novel approach of combining performance metrics, legislative drive cycles and distance-based drive cycles, vehicle simulation is used to better understand the design space across different scenarios. This understanding allows the development of powertrains that are optimised for an autonomous vehicle's expected usage pattern which in turn makes it possible to state their potential efficiency gains.
The key results provided are the vehicle energy consumption, vehicle range, vehicle availability, total time taken to complete a drive cycle and the Total Cost of Ownership (TCO). The paper also gives indications of the powertrain mass and cost benefit analysis attributable to a BEV passenger car powertrain size that is optimised for Level 5 autonomy.
