Driver on a country road
Health aspects are becoming increasingly important in the digitally connected safety services that are being developed for vehicles. This is made possible by smart sensors and artificial intelligence - and also by an increased interest in such services.​ Photo: Johan Karlsson, Autoliv​​​​​​​​​

A technology that alerts if you doze at the wheel

Researchers at Chalmers and Autoliv are developing an IT platform that, in the vehicle setting, recognises the signs of a driver getting sleepy. In the beginning of next year, the system will be tested by 50 drivers on public roads in Gothenburg.
“Sleepiness is the cause of many traffic accidents, and by using new technology we can help drivers to refrain from driving when they are not completely alert”, says Stefan Candefjord, Assistant Professor at the Department of Electrical Engineering, Chalmers University of Technology.

The aim is to measure, analyse and communicate physiological signals in a vehicle. The project is called COPE, Connected Occupant Physiological Evaluation, and is run jointly by researchers from Chalmers and Autoliv.

When a person becomes drowsy and sleepy, the heart rate and breathing are affected, among other things. By integrating sensors in the car interior, for example in the seat belt and the steering wheel, the variations in heart rate and respiration can be monitored in real time. Smartwatches and bracelets with built-in sensors, worn by the driver, can also be used to register physiological signals.

Promotes road traffic safety and health 
“Traffic safety is in focus, but also to handle sudden acute illness, to follow-up on chronic health conditions or treatments, and to take the adequate actions in the event of traffic accidents”, says Bengt Arne Sjöqvist, Professor of Practice Emeritus at Chalmers, and since many years active in the area of digital health.

“In the digitally connected safety services that we at Autoliv are developing, the health aspects are becoming increasingly important”, says Pernilla Arnell, responsible for business development within Autoliv. “Making the driver aware of health conditions that may affect driving is a natural next step. Our surveys show that there is a positive attitude – ​ not least among middle-aged women – towards various aids that promote safe driving.”

The technological advances are governed in that direction also via legal requirements and scoreboard assessments of safety functions at the EU level, which regulates and rewards the implementation of sleep detection systems in new cars.

Shifting focus from the vehicle to the driver
“As technology for autonomous steering and assisted driving becomes more common in cars, the drivers will probably get tired faster, especially at night, when the task to drive is less stimulating”, says Johan Karlsson, Senior Research Engineer at Autoliv Research. “That makes systems for reliable detection of sleepiness even more important.”

Today, many cars already have some kind of built-in sleepiness detection that is based on the driver's performance by monitoring how the vehicle moves on the road. However, if the driver's unfocused driving is corrected by the car's driver assistance system, this loses its significance as a measure of sleepiness. In the future, there will be other, more direct ways of measuring drowsiness, where the focus is set on the driver's physiology and not on determining the driver's performance through the the behaviour of the car.

A hardware has been developed by Autoliv, which connects sensors in the car with online services. This also makes it possible, for example, to use information from fitness watches or other smart watches, to transfer your driver profile to vehicles in carpools, and to take advantage of individual adaptation of sleepiness detection.

“The results are interpreted by an algorithm trained by artificial intelligence, which we at Chalmers have developed in collaboration with Autoliv and VTI”, says Stefan Candefjord. “The system recognises the signs of a person going into drowsiness, thus having an impaired ability as a driver. Data which is collected during the drive can be shared to the cloud, and of course also be used by systems in the car that ensure that the driver, if possible, focuses his or her attention again or gets a recommendation to take a break from driving.”

“We are currently evaluating whether unobtrusive sensors can provide as valuable information as medical-grade ECG sensors”, says Ke Lu, Postdoctoral researcher, working in the project group at Chalmers.

Driver on a country road


Test drivers will receive a small black box to put on top of the instrument panel, which collects data on the driver's respiration and heart rate. The car in the picture is equipped with a steering wheel with built-in ECG sensors.

Testing on its way
In the beginning of next year, the research will be brought out on public roads to test the system on a larger scale and to collect data during realistic conditions.

“We plan to use a test fleet of about 50 vehicles”, says Johan Karlsson. “In the first phase, the drivers will be asked at regular intervals to estimate their level of sleepiness on a touch screen. In parallel, the physiological signals are registered by heart rate monitors, and for some drivers also by prototype sensors in the steering wheel and seat belt.”

Data collected is used to further calibrate the algorithms. They should be able to detect if the driver is drowsy, and to some extent also predict how the level of sleepiness will progress in the next few minutes.

The IT platform that is being developed within the COPE project is considered to be a central part for other applications and research projects, for example to add important information to smart alarm handling systems for increased traffic safety within the TEAPaN project.

Text: Yvonne Jonsson
Photo: Autoliv
Johan Karlsson in a vehicle simulator


Johan Karlsson, researcher at Autoliv, in the simulator where new sensors that are able to register the driver's respiration and heart rate will be tested, for example integrated in the seat belt.

More about the COPE project
Connected Occupant Physiological Evaluation, COPE, is a two-year research project that aims to develop and test smart monitoring of health data in real time with a focus on sleep detection in drivers. Chalmers Transport Area of Advance​ and Autoliv are funding the research, that is conducted in connection with SAFER, the Vehicle and Traffic Safety Centre at Chalmers.


Are you looking for a master’s degree project and want to contribute to this research?
Do your thesis on smart seat belts in collaboration with Chalmers and Autoliv: Monitoring driver's respiration and drowsiness using smart seatbelt



For more information about Chalmers’ research on digital health, contact:
Stefan Candefjord, Assistant Professor in the Biomedical electromagnetics research group, Department of Electrical Engineering, stefan.candefjord@chalmers.se​
Bengt Arne Sjöqvist, Professor of Practice Emeritus in the Biomedical signals and systems research group, Department of Electrical Engineering, and Head of Business & Strategy at Prehospital ICT Arena (PICTA) at Lindholmen Science Park, bengt.arne.sjoqvist@chalmers.se
Anna Sjörs Dahlman​, Adjunct Associate Professor at the Department of Electrical Engineering and researcher at the Swedish National Road and Transport Research Institute (VTI), anna.dahlman@vti.se
Ke Lu, Postdoctoral researcher in the Biomedical signals and systems research group, Department of Electrical Engineering, ke.lu@chalmers.se
Anna Bakidou​, PhD student in the Biomedical signals and systems research group, Department of Electrical Engineering, and the University of Borås, bakidou@chalmers.se

For more information about Autoliv’s research, contact:
Johan Karlsson, Senior Research Engineer, Human Factors, Autoliv Research, johan.g.karlsson@autoliv.com
Pernilla Arnell, Director Business Development Sales, pernilla.arnell@autoliv.com
Per Gustafsson, Group Manager Innovation and Testing, per.gustafsson@autoliv.com


Published: Fri 06 Nov 2020.