Impacts of Connected and Autonomous Vehicles on Traffic Flow - A Study on U K Traffic Networks

Impacts of Connected and Autonomous Vehicles on Traffic Flow - A Study on U K Traffic Networks


David Williams, Atkins Ltd, Xucheng Li, Atkins Ltd


This study investigates the potential impact of connected and autonomous vehicles on the UK road network. The focus is on potential changes to longitudinal and lateral behaviour, and how this may ultimately impact traffic network performance.


The opportunities presented by the emergence of connected and autonomous vehicles (CAVs) are vast in scope. At the heart of this is likely to be a fundamental change in mobility; how and why people travel. Change on such a scale will inevitably take a substantial amount of time to achieve.

In the intervening period, the increasing penetration of connected and autonomous technologies in the vehicle fleet will likely influence traffic flow and ultimately road network capacity. The drivers of this change are microscopic in nature, resulting from differences in longitudinal, lateral and accelerating behaviour of vehicles.

Whilst much previous work has investigated a step-change in vehicle capability (akin to 100% fleet penetration of CAVs), there is a need to understand how incremental changes to the fleet may impact traffic flow and capacity. Atkins were commissioned by the UK Department for Transport to provide a robust evidence of the potential impacts of CAVs on the UK road network. This paper describes the methodology employed and the results from this study.

A traffic microsimulation platform offers the opportunity to influence traditional models of vehicle dynamics and driver behaviour so as to proxy the potential impacts of CAVs. In this study, PTV VISSIM 8 microsimulation software is adopted.

This work has sought to understand the impact of changes to longitudinal behaviour, lateral behaviour and an enhanced provision of information on traffic networks in the UK. An extended section of the Strategic Road Network (highway) is modelled, allowing behaviour and performance to be evaluated around:

- Free-flow links and interchanges;
- Merge and diverge at junctions;
- Signalised junctions and stop-lines; and,
- Different prevailing conditions of congestion.

Building on the foundations of previous studies, the modelling platform used here has been enhanced through use of the VISSIM COM interface. This enables CAVs to adopt different behaviour sets depending upon the situation.

Given the uncertainty surrounding the behaviour and uptake of CAVs, a “what-if?” analysis is undertaken. Systematic variation in both the capability and penetration of the connected and autonomous vehicle fleet allows the range of potential effects to be quantified. Results demonstrate that the potential for improvements to traffic flow regimes are highly dependent on penetration and performance, and that there is the potential for disbenefits in the short term.

In summary, by analysing a range of potential futures for CAVs on the UK road network, this paper aims to inform policy makers of the potential benefits and disbenefits and identify priorities for future research.


Association for European Transport