Autonomous driving: Q&A with Chris Hoyle, Technical Director at rFpro
We are continually told that driverless cars are tomorrow's world. Already the idea that they will feature at the 2020 Olympic Games is a given. In the meantime, an engineering team at rFpro have been looking at the ways in which to test these vehicles accurately in a safe, virtual environment using a simulator. To find out more, Matthew Beecham spoke to Chris Hoyle, Technical Director at rFpro.
How is the simulator sector helping to bring autonomous vehicles to market?
The automotive industry's ambitious jump to autonomous driving is huge, both in terms of technology and consumers' ideology. While the technology to introduce autonomous driving is already in existence, a long and thorough test schedule stands between manufacturer implementation and end-user uptake. Simulator technology plays a crucial role during this transition, allowing vehicle manufacturers to test autonomous vehicles under the most challenging conditions, facing every eventuality, freak scenarios and in complete safety.
The benefit of the repeatable nature of simulator tests is that data can be collated that is not affected by third parties; changes in weather, heightened driver anxiety or the intervention of other vehicles, for instance. Originally developed for use in Formula 1, the latest simulator technology by rFpro reacts 100 milliseconds faster than typical 3D engineering graphics - which translates to a car length at motorway speeds - and can offer test conditions providing all road and seasonal conditions, temperatures and light. Such pinpoint simulation allows vehicle manufacturers to manage the transition between 'driver in control' to 'car in control' gradually, accurately assessing the required need for communication between machine and human.
Are vehicle manufacturers working to address customers' concerns about relinquishing control to autonomous vehicles?
Reassuring the next generation autonomous vehicle 'drivers' over their decision to relinquish control to the car is a crucial step for initiating commercial uptake of these technologies, and it is an area in which vehicle manufacturers are heavily investing. Ironically, driver in the Loop (DIL) simulators can play an integral part in gradually removing drivers from the loop and investigating the reaction of autonomous passengers: how can vehicle manufacturers reduce the level of inappropriate driver intervention - grabbing the wheel when approaching a perceived danger, for example - and inspire confidence in the vehicle's decision making processes. Manufacturers are able to examine the best way for vehicles to indicate their intentions to the driver, helping to allay misplaced uncertainty and increase 'driver' trust. Autonomous vehicle control systems will be far more spatially aware and possess the ability to react to sudden peril far quicker than humans, but the successful communication and reassurance to inspire trust in this capability is crucial.
Of course, public testing of autonomous vehicles has to become a reality before the technology is rolled out for the first time commercially, but simulator technology can give vehicle manufacturers a head start in the development of driver interface long before a vehicle needs to be tested on the road. Professional test drivers getting to grips with autonomous technology does not provide an accurate representation of how consumers will view the prospect of handing back control, who require greater reassurance.
To combat these preconceptions, vehicle manufacturers can simulate a virtual world environment where numerous 'real' drives are experienced on the same road networks, allowing 'drivers' to experience on-road interaction between autonomous vehicles and develop tailored control strategies that best soothe consumer fears of an autonomously-run road network.
Presumably a 'driverless' future makes for a safer road network - do you think this will be a key selling point of autonomous vehicles?
If we look at aviation, where autopilot systems have been successfully employed for the last century, statistics from the Civil Aviation Authority (CAA) state that human error is the main cause of aircraft accidents. Therefore, this suggests that the removal of humans from the decision making loop improves safety, and there is no reason why that should not transfer to the automotive industry. However, road vehicle occupant and pedestrian safety is now so impressive that this alone is unlikely to provide sufficient incentive for drivers to willingly relinquish control.
Instead, convenience and functionality provide far greater enticement. Who wouldn't want to hand responsibility for the daily, stressful hack through commuter traffic and down bland highways over to the vehicle itself?
How would utilising simulator technology in this way impact on development costs?
The remainder of this interview is available on just-auto's QUBE Global light vehicle safety systems market- forecasts to 2030