During the months of June and July, VENTURER undertook the second tranche of experiments. Trial 2, held at the Bristol Robotics Laboratory, involved examining human reactions to the decisions made by an autonomous vehicle at junctions and while interacting with conventional vehicles.
Participants engaged in a series of repeated scenarios in both the Wildcat and VENTURER Simulator, and were asked to rate their ability to trust the decisions made by the vehicle.
The scenarios were designed to mirror the experiences of an everyday driver on UK roads and included overtaking a parked vehicle with a vehicle oncoming and turning left and right into and out of junctions while detecting passing vehicles.
It is expected that the findings from Trial 2 will help to inform future research around Connected and Autonomous Vehicles’ (CAVs) abilities to integrate with other traffic on UK roads.
Two of the participants that took part in the Trial 2 experiments were Kurt, who has 28 years driving experience and Megan, who has just 5 years of experience.
Kurt, when asked to reflect on his time in the Wildcat vehicle, stated that “what was interesting was that I trusted the car even though it makes different decisions to what I would have made if I was driving” and importantly, he never felt that the vehicle’s decisions were unsafe. Regarding the interactions with other vehicles, he felt that the car was “perfectly fine with cars crossing its path”.
Although he could see how the Wildcat and VENTURER simulator experiments were comparable, he felt more at ease in the Wildcat than the simulator as he found it easier to concentrate on the vehicle’s movements due to the sensations of the engine and the scenarios felt more real-life.
Megan also felt safe in both the simulator and the road-vehicle claiming that “at no point did I think it was going to crash or it was going to be unsafe”. She felt that although the vehicle was often “cautious around junctions”, she didn’t feel nervous at all and thought that the CAV “made decisions that a normal driver would have in certain situations”.
When asked to consider the biggest barriers to wide scale CAV adoption, Kurt and Megan had differing views. Through Kurt’s exploration of CAV trials in the news and through technology blogs he is aware that projects like VENTURER and the existence of technologies such as adaptive cruise control are proving that technological progress is occurring at a rapid rate. As such, Kurt felt that “human attitudes and behaviours towards driverless cars are the biggest barrier” currently preventing wide scale adoption of CAVs rather than the technology enabling them. Megan, on the other hand, stated that she felt the most challenging aspect would be stepping-up the technology’s abilities from functioning in a controlled environment to functioning in the real road.
Neither Megan or Kurt felt that CAVs are yet fully equipped to anticipate and react to the unpredictable driving behaviours that human drivers exhibit on UK roads and can see why projects like VENTURER are so important in improving this.
Kurt also suggested that pedestrian attitudes towards CAVs might be counter intuitive as their greater confidence when stepping into the road in front of autonomous vehicles may lead to increased jaywalking. While this may be positive for the safety of pedestrians and other road users, this could result in delays on the road network and slower journeys. Despite these issues, he agrees that it may be possible to mitigate against the issues surrounding pedestrians taking advantage of driverless vehicles by making CAVs indistinguishable from normal cars.
Both Kurt and Megan identified an autonomous vehicle’s ability to interact with pedestrians as a concern. This is something that VENTURER will explore extensively during Trial 3 as it is a crucial issue that needs to be assessed before driverless cars can be safely deployed onto UK roads. Furthermore, CAV technology’s ability to cope in real world will also be put to the test in Trial 3 when VENTURER will demonstrate an autonomous vehicle in more complex scenarios in the Bristol and South Gloucestershire region.