Transport modelling and management for the transition to autonomous driving

Professor Zhang, Fangni   (Principal Investigator (PI))

24

2021-01-01

150000

Transport modelling and management for the transition to autonomous driving

Autonomous vehicles, Behavioural shifts, Infrastructure planning, Service design optimisation, Shared transport, Transport network modelling

Transportation

Engineering (E)

202009185002

Seed Fund for Basic Research for New Staff

2020

Completed

This project aims to improve the modelling and management tools to address emerging transport challenges brought by autonomous vehicles. With the rapid development of self-driving technologies, autonomous vehicles are becoming an integral part of the future transport landscape. This is triggering evolutions in the ways people travel and live. While driving-assisted vehicles have been penetrating the car market at an increasing rate worldwide, years or even decades are expected before fully autonomous driverless vehicles can be ready to completely replace traditional vehicles. In the near future, more prevalent is the situation where the road traffic is a mixture of both traditional vehicles and multiple levels of autonomous vehicles, among which the extent a car takes over responsibilities from its driver varies. This gives rise to a critical need to understand travel behaviour and to efficiently plan and manage transport infrastructures in the transition period. This project intends to address this knowledge deficit by developing new analytical methodologies characterising traveller responses to multiple autonomous driving options, transport system performance with mixed traffic, and transport operator management strategies, as well as the complex interactions among traveller, infrastructure and operator. The outcome of this project expects to inform policy and decision makings for the efficiency, resilience, safety, and sustainability enhancements of urban systems transitioning to the era of autonomous driving. This project aims to generate insights for effectively modelling and efficiently planning, managing and operating transport systems with the coexistence of traditional vehicles and various levels of autonomous vehicles. The project investigates the interactions among travellers, transport infrastructure operators, and transport planners in the progress of autonomy. From the traveller’s perspective, the tradeoffs among travel time, driving effort, riding comfort, energy consumption, reliability, safety and privacy vary across vehicle types (traditional or autonomous, private or shared) and driving modes (higher or lower automation level). This project intends to identify how these tradeoffs would affect traveller’s travel choices of not only travel mode, departure time and route, but also the driving mode. The former three types of choices have been present for long; however, the driving mode choice is recently emerged, which involves the choice between human-driving and autonomous-driving, as well as the choice among various levels of automation. While there is a lack of understanding of this new complication, this project aims to envisage its system-wide impacts on travel behaviour, on traffic congestion, and on system reliability. Apart from the mentioned short-term travel behavioural shifts, household’s longer-term decisions on car ownership and housing relocation also play a role in reshaping human mobility patterns in urban cities. Household car ownership, in this context, refers to the combined decisions on whether to possess private cars, how many cars to have, and more importantly, what automation level to choose. This project intends to anatomise these decisions and relate them with travel choices and housing location choices. When commute with autonomous vehicles becomes the productive and entertaining extension of a variety of other activities, the pressure for commuting delay may substantially change, which may in turn affect the preference for housing location and the demand for travel, forming a feedback loop. This project aims to reproduce human mobility patterns under extended influences of autonomous vehicles in the long run. Considering that transport infrastructure generally has a long lifespan, this project also intends to incorporate the anticipated future needs of autonomous vehicles into the transport and urban planning. A smart design of transport systems involves efficient use and allocation of the limited urban space in the mixed traffic environment over both spatial and temporal horizons. The project aims to understand how transport planner and infrastructure operator should cope with multiple vehicle types, plan new automated roads and/or manage existing networks. To these ends, the system-wide effects of expanding and distributing the road space will be analysed taking traveller’s short-run and long-run responses into account. Specifically, the objectives of this project are: (i) To establish new equilibrium formulations characterising traveller’s behaviour in terms of travel choice (e.g., travel mode, departure time, and route) and driving mode choice (e.g., human or autonomous driving, level of automation); to uncover the interactions among different groups of travellers (e.g., traditional vehicle user versus autonomous vehicle user, higher automation level user versus lower automation user, and private car versus public transport user); and to pinpoint effects of autonomous driving on dynamic travel behavioural shift, on traffic congestion, and travel time reliability. (ii) To expand the modelling horizon to incorporate household’s car ownership and housing relocation decisions; and to identify how these factors will evolve with the adoption of autonomous vehicles, and how the evolutions will reshape the traffic dynamics in the long run and so forth the transport system performance. (iii) To develop planning and management strategies for transport planner and infrastructure manager respectively; to devise analytical tools to inform optimal location for new autonomous driving roads and how they should be connected to the existing network; and to identify efficient space and infrastructure management strategies for urban and transport systems.