Can driving simulators help advance the connected car?
Picture credit: Ansible Motion
Driver in the Loop (DIL) simulation has long been considered a powerful tool for both vehicle and component development. For traditional vehicle development work – such as ride and handling, steering, and powertrain – driver feedback is essential, offering subjective information that no data logger can deliver.
Simulators place real people into direct contact with experimental systems, and this is arguably the main reason why so many car makers have invested in them. As simulators become more widely accessible to engineers and grow ever more computationally powerful (our own Delta Series simulators are typically powered by an array of 16 5GHz PCs), their performance levels are rising to the point where they also can be used for validating Advanced Driver Assistance Systems (ADAS), and various connected vehicle systems.
As simulators grow ever more computationally powerful, their performance is rising to the point where they can be used for connected vehicle systems
Validation of such systems within a virtual world becomes more effective with ‘engineering-class’ simulators, those that are able to make a driver believe he or she is interacting with an actual car on an actual road, through playing ‘tricks’ on the multiple human senses responsible for optical (vision), haptic (touch), audio (sound) and vestibular (movement) reckoning.. The move away from traditional hexapod-based “pointing” simulators to stratiform-based dynamic simulators , such as that found in Ansible Motion’s Delta series, is but one example of the emerging technologies that enable skilled drivers to conduct meaningful virtual test drives, crucially those that can be validated against in-car data from driving sessions in real cars.
With auto makers focused on exploring the myriad technological possibilities for the connected car, a less-than-capable simulator can unfortunately lead to distraction. An engineering-class simulator coupled with a multi-layer validation process will ensure the delivery of trusted information. This disciplined approach can be a game changer in terms of efficiency because, once the DIL simulator is in place as a tool, a vast number of human interaction experiments can be conducted in a very short period of time.
There are three fundamental layers to this validation process. The first layer is objective, wherein a comparison of data between the simulator and a real vehicle draws out any exaggerated or understated driver control behaviours. The second layer is subjective, wherein qualified driver feedback is used to analyse whether specific experiences in the simulator match expectations. Once the objective and subjective layer corrections are in place, then finally there is the third layer, which is the ultimate test: carefully developed vehicle and simulator tuning changes are deliberately introduced to see if the driver feedback matches the known changes. Only once this has been achieved can the actual vehicle engineering work commence.
Once confidence in the validation process has been established, auto makers and their suppliers can use DIL simulation experiments to reduce the validation time for advanced driver interaction systems and autonomous handover protocols. What once took ten days of in-car testing for a single prototype subsystem can now be completed within three days when the test matrix has been vetted in advance in a capable DIL simulator lab.
Correctly configured Driver in the Loop simulators are safe and consistent laboratory tools for putting human drivers into unexpected or emergency situations while interacting with on-board vehicle systems
Correctly configured and validated DIL simulators are safe and consistent laboratory tools for putting human drivers into unexpected or emergency situations while interacting with on-board vehicle systems. Testing active safety systems such as forward collision warning, collision mitigation braking, pedestrian avoidance, intelligent interaction systems and cooperative driving systems are but a few application examples. In the realm of autonomous vehicle development, there are just as many applications, if not more since the introduction of autonomous vehicles will put people into completely new situations, without historical precedent, requiring safe and scientific investigations. Where else but in a DIL simulator lab might such studies take place?
We might assert with some confidence that connected car developments, in all their forms, will necessitate that more and more testing and validation be undertaken in the virtual world. Certainly a part of this can be off-line simulation. However, as long as cars remain consumer products, it will be important for people to be able to interact with them pleasurably. As such, it is likely that Driver-in-the-Loop simulation will continue to play a vital role in bringing cars to market in the decades to come.
If you’re interested in finding out more about 5G, head over to the 5G conference and exhibition page for more 5G insights.