Articles
Dr Stewart Birrell, assistant professor of experiential engineering at WMG, University of Warwick
Connected and autonomous vehicles (CAV) are coming to market whether the majority of drivers want them or not. CAV are being driven by a ‘technology push’ rather than a ‘consumer pull’. The technology is expected to be widespread by 2020. However, testing and validation of the robustness and trustworthiness of automated systems has lagged behind, which is why Warwick Manufacturing Group’s 3xD Simulator for Intelligent Vehicles was developed.,
When we started developing the 3xD Simulator for Intelligent Vehicles, we didn’t want to simply replicate another simulator facility. My colleague, Gunny Dhadyalla, principal engineer at WMG, and I set about developing a simulator that would be a world first. We identified three key components: the ability to drive in any vehicle, either production or prototype; to house it inside a Faraday cage, so we can emulate radio frequency signals and make the car think that it is receiving an actual GPS signal or a wi-fi hotspot; and to provide a highly realistic driving environment.
Our research is focused at the applied level to facilitate technology transfer for industry. We are involved in two industry-led research projects. The first is with RDM Group, the company behind the low-speed autonomous pods that will be on the streets of Milton Keynes next year. The Intact project aims to speed up the validation and safety case using the simulator to replicate dangerous scenarios that the pod could encounter in the real world.
The second project is UK CITE, which aims to create a road network in the Midlands where connected cars can be trialled. Here we are evaluating the cyber security of connected cars, again using the simulator, and trying to understand future business models of vehicle ownership.
Aligned with my research interests in experiential engineering is the issue of user trust and misuse of autonomous features in vehicles. At WMG we are researching the so-called ‘calibration of trust’ in autonomous vehicles. Early adopters may trust the technology too much, whereas most will have very low levels of trust, preventing the use of automated safety features or leading them to say they won’t buy a self-driving car.
Another area of interest for me is driver state monitoring (DSM), where the driver is monitored when inside the car, to determine if they are sleepy or not paying attention to the road. With DSM you can personalise warnings, by determining if mirrors have been checked by monitoring head position, for example. These systems will be a common feature in premium vehicles within a few years, and will use sensors and cameras. But we can also use the ‘state’ data that smartphones and wearable electronics such as fitness trackers provide instead of expensive embedded sensors to make DSM more accessible for all drivers.
We are also working on emulating radar in the simulator, to test advanced systems such as adaptive cruise control.
Driving simulators have primarily focused on how the driver reacts to certain situations, not how the vehicle’s technology deals with these scenarios. The future simulation environment will contribute more to virtual prototyping and validation by evaluating new systems in the virtual environment and bridge the gap between benchtop development and real-world testing.