PhD Awarded for thesis on autonomous heavy vehicles: Garrett Bray

Garrett Bray has successfully completed and defended his PhD thesis entitled ‘Autonomous Goods Vehicles: Implications for fleet operating models’, (University of Cambridge, April, 2022).  The thesis investigates the cost and carbon benefits that are likely to be available through use of autonomous delivery vehicles for a variety of logistics activities. The abstract is provided below.

Abstract

The potential impact of autonomous vehicles (AVs) is becoming increasing relevant as a result of developments in technology, industry investment and regulations such that there appears to be a reasonable probability of on-road deployment of autonomous vehicles at scale in some geographies this decade. Road freight constitutes a significant sector of the transport industry and proponents believe that trucking applications present a greater near-term opportunity for feasible deployment of Avs than passenger mobility. Despite the likely near-term deployment, the research and understanding of potential impacts of autonomous trucking is less developed, with few efforts to date involving detailed quantified analysis and fewer still examining impacts beyond the first-order changes to the total cost of ownership. This thesis explores this gap in the literature to examine potential changes to operating models and the consequent economic and environmental impacts that may be enabled by the removal of the human driver.

Three operating model changes are examined using a mix of parameterised analysis and case studies: (I) speed strategy; (II) selection of vehicle size and extent of multi-drop deliveries; and (III) off-peak deployment. In Chapter 2, target highway speed adjustments are examined using a parameterised analysis of the trade-offs of fuel, vehicle, driver and freight value of cargo time. A validated fuel consumption model is used to compare fuel consumption for different speed strategies across idealised highway drive-cycles and vehicle scenarios. The chapter concludes that the adoption of lower speeds presents an indirect savings opportunity in addition to savings associated with the removal of the driver. In Chapter 3, a set of increasingly complex applications are examined on the selection of vehicle size and extent of multi-drop deliveries, ultimately incorporating a unique vehicle routing problem formulation for two multi-destination applications. The chapter concludes that the use of smaller vehicles and fewer deliveries per journey could produce incremental savings, dependent on the application. In Chapter 4, on off-peak deployment, a modelling technique is developed to link fuel consumption with traffic congestion and time of day. It uses journey duration estimates from Google Maps, the segmentation of journeys by road speed limits using HereMaps and fuel consumption as a function of congestion using the HBEFA traffic situation fuel consumption database for the different road types. A unique form of the time dependent vehicle scheduling problem is created. The chapter concludes that off peak hour deployments present the potential for incremental value creation, derived largely from reduced fleet size requirements, in scenarios where human drivers are paid shift loading for off-peak hours. In Chapter 5, combined changes are also examined and it is concluded that reduced highway speeds and off-peak deployment present the greatest opportunity for combination. This analysis was supported with a modelling technique developed to enable consideration of reduced vehicle speeds as a decision variable in conjunction with the impact of congestion for different times of day.


In summary, this research shows that incremental cost savings and environmental benefits can be achieved through modifying operating models for autonomous trucks in addition to the first order impacts of removing the cost of the human driver.