AVs: demonstrating caution could be the key to public trust

In a second government-backed Venturer autonomous vehicle trial the levels of trust between the car and passenger were measured.

Autonomous vehicles: operating cautiously increases trust
Autonomous vehicles: operating cautiously increases trust
  • Updated: 23 January 2018
  • Author: Imogen Weight, Graduate Transport Planner, Strategy and Business Case, Atkins

Demonstrating that autonomous vehicles (AVs) are able to safely navigate different highway situations including interactions with other vehicles is a crucial aspect of ensuring they will be safe for deployment on UK roads.

Venturer’s second trial was dedicated to examining how an AV performs when interacting with other vehicles on the highway and the extent to which participants trusted the AV’s decisions during the manoeuvres it performed.

The trial, based on 10 typical highway scenarios, was conducted using both the Wildcat and the Venturer simulator. A total of 46 participants engaged in the experiments last summer and were asked to rate their trust in the AV’s decision following each scenario.

The Venturer researchers found that participants generally always indicated high levels of trust towards the AV with ratings exceeding 7 out of 10 (where 10 represented complete trust) in virtually all scenarios.

In keeping with predictions, participants’ overall trust ratings were higher for the Venturer simulator than the Wildcat, perhaps reflecting a perceptual evaluation of trust when being driven autonomously in a real vehicle with a real cost to the person if things went wrong.In the simulator there is less personal investment required.

Participants were also required to complete a series of psychometric tests to provide information on individual differences in factors such as driving experience, personality, and trust in technology and automation. Interestingly, preliminary analysis of this data showed that there was no significant correlation between trust scores and age.

Following the main participant experiments of Trial 2, 41 new participants took part in a series of simulator only experiments. These experiments were included to form a comparison of how trust scores varied depending on the programmed behaviour of the AV.

Overall analysis of these experiments indicated a need for AVs to operate more cautiously than the average driver to foster public trust. This could have significant implications for the future UK road network, for example, if AVs were to drive more cautiously than the average human driver they may create a traffic calming effect on the network.

This may result in safety benefits in conditions with mixed AV and manually driven vehicles. Furthermore, there might also be congestion benefits in certain traffic flow conditions as a result of smoother traffic flow from the greater level of consistency in gap acceptance behaviour.

To fully understand the implications of the use of different AV modes on user acceptance, road safety, congestion and manufacturing standards, further research and development will be required.

Venturer and Atkins, the project’s lead partner, support this and anticipate working with government and industry to develop standards around AV operating modes to underpin a robust legal and insurance model for the production and operation of AVs and their safe deployment onto UK roads.

You can read more about the Venturer project and find the full results from Trial 2 here: www.venturer-cars.com

You can also connect with us on Twitter: @Venturer_cars