Simulator shows how humans interact with autonomous cars

Assessing the unexpected
With the driver assistance systems market set to grow to $70bn by 2024, fuelled by vehicle manufacturers pushing toward increasing levels of autonomy to address emerging legislation, the issue of how real people might react to a car receiving more notifications - or even taking control - is one that car makers are investigating seriously.

“Car makers are introducing more driver assistance technologies, but their level and method of intervention differs by car brand,” said Kia Cammaerts, Founder and Director of Ansible Motion.

“If a car does something unexpected, we are able to test what the driver and occupant reactions will be in our simulator laboratory, well in advance of cutting any metal. Our latest simulator enables car manufacturers to design better and safer vehicles and assess many proposed technologies early in the design cycle.”

New tricks
To create the most immersive human simulation experiences - and therefore the most meaningful pre-validation results - Ansible Motion’s simulator lab in Hethel, Norfolk, has now added a number of new features, such as new cabin environments that reflect OEM styling and human interaction features and new software connectivity that allows deeper environment and sensor simulation, coupled with Ansible Motion’s proprietary motion, vision, and audio environment that ‘tricks’ drivers and occupants into believing they are experiencing a real vehicle and its ADAS or autonomous technologies.

This special type of driving simulator technology, pioneered by Ansible Motion, is far-removed from gaming 'simulation' endeavours, and has been trusted for nearly a decade by top vehicle constructors in the US, Europe and Japan (Ford, GM and Honda, to name a few) has now been updated extensively to offer a realistic environment for evaluation of the newest on-board vehicle systems and concepts.

“We aim to deliver compelling experiences to connect real people to the world of simulation.  This method of providing virtual ride and drive experiences has proven to be highly effective for vehicle constructors as they trial their new concepts,” claimed Cammaerts.

Cutting evaluation times from hundreds of years to months
With the ability to create and explore an incredible number of scenarios in a short amount of time, Ansible Motion’s simulator means engineers can conduct experimental variations that might consume a hundred years’ of testing time in the real world, within a few months.

Examples include the validation of Autonomous Emergency Braking (AEB) systems that rely upon multiple sensor feeds and vehicle piloting logic algorithms to respond (in some cases, faster that human response capability!) to various situations such as traffic and pedestrian intrusions. 

The need to validate international drivers’ expectations
Other system validation examples include lane departure warnings and assistance, intelligent speed adaption and driver monitoring for drowsiness and distraction. Cammaerts cites a recent example of how drivers in the Chinese market expect different intervention cues for lane departure warning compared to their US counterparts.

“There are cultural differences and expectations to respond to audible or visual warnings. Validating this in our simulator prevents frustration, dissatisfaction or confusion when vehicles deployed in different markets are required to interact in critical situations.”

“With the burgeoning need to validate more and more driver assistance systems and autonomous functions, the number of possible scenarios grows every day,” added Cammaerts.

“It’s, of course, impossible to validate every situation on a proving ground test track or in the real world: there simply isn’t time. And in the case of real world testing, well, some experiments may be quite dangerous, putting people and equipment at risk. Simulation technologies enable car makers to validate their required cases in a controlled lab environment and do so safely and quite efficiently.”