Smarter and safer automated driving
Hi-tech driver assistance systems being developed by EU-funded researchers aim to perfect the link between human and machine in self-driving cars. This could help prevent accidents and fatalities on Europe's roads.
© pavelvinnik #243439695, 2019 source: stock.adobe.com
Automated driving is soon set to change the way we drive, but its success depends in large part on how effectively this cutting-edge technology links to the living, breathing person behind the steering wheel.
If that link can be perfected, then it could herald a future in which human errors are compensated for by intelligent software, leading to safer trips and more efficient road usage for all vehicle types.
Enter the EU-funded ADASANDME project. Its partners are working together to take everything we know about automated driving and combine that with emerging knowledge of how to predict and detect key factors affecting a drivers state such as fatigue, stress, inattention, fear and fainting in various situations.
By developing advanced driver assistance systems, ADASANDMEs work aims to ensure that vehicles can switch safely and reliably back and forth between human drivers and automation modes.
Regardless of what vehicle you drive, youll suffer from being impaired now and then, says ADASANDME project coordinator Anna Anund of VTI, the Swedish National Road and Transport Institute. But our challenge is to only return control to a driver who has been under automation if that driver is fit to take over. At the same time, we need to make sure the driver benefits from this to make sure that drivers are more rested and less stressed.
Trucks, bikes and automobiles
With the analysis of drivers needs and the current state of the technologies now complete, the remainder of the project will see the ADASANDME team focusing on seven key areas to identify, clarify and organise the requirements of the systems. These areas are:
The researchers are developing robust detection and prediction algorithms to monitor different driver conditions, using both existing and novel sensing technologies that will take traffic and weather conditions into account. They are also working to personalise these systems so that they can be tailored to the physiology of the driver and his or her typical driving behaviour.
Ultimately, the newly developed human-machine interfaces and automation systems will be evaluated with a range of drivers with real vehicles, on real roads, to ensure they can accurately detect when the driver is unable to drive safely, take control, and avoid danger.
Towards the end of the project, the team will be checking that the advanced driver assistance systems will be able to accurately recognise the drivers condition and offer clear information to the driver. The researchers will also be investigating the extent to which drivers trust and accept the system, how they react to system warnings and suggestions, and collecting their opinions on its use.
There are several innovations in our project, ranging from new driver state detection algorithms based on physiological signals and eye trackers adapted for automation, to new, more effective human-machine interfaces, says Anund.
There is plenty of market potential for technologies developed over the course of the project. For example, sensor and algorithm developers are working on a number of innovations that are likely to be strong candidates for commercialisation in the near future.