The strength of the BMW M4 Coupé is evident in every detail.
Gothenburg, Sweden - Volvo has combined the sensor technology used in self-parking systems with facial recognition software to create a car that literally keeps an eye on its driver, gets to know him - or her - learns when they're tired, distracted or driving badly and warns them about it.
And if that sounds a little too 'nanny state' for you, it's part of the company's commitment that by 2020 nobody should get killed or badly hurt in a new Volvo.
A sensor on the dashboard monitors the position and angle of the driver's head, the direction in which they're looking and how open their eyes are.
From that it's possible to develop systems that can gauge the driver's condition fairly precisely and enable the car to adjust accordingly.
And by connecting the monitor to the car's existing driver aids, such as lane-keeping assist and adaptive cruise control, the system can ensure that the car doesn't stray out of its lane, get too close to the car in front or take out a pedestrian while the driver is changing CD's, sending text messages or yawning - and it'll sound a particularly nasty beeper if they fall asleep.
Driver support project leader Per Landfors added: "Once the car is able to detect whether its driver is paying attention, safety systems can be adapted more effectively.
"For example, the car's support systems can be activated later on if the driver is focused, and earlier if the driver's attention is directed elsewhere." explains.
HOW DOES IT WORK?
Small LED's on the dashboard light up the driver's face with infrared light, which the sensor is tuned to see but is just outside the wavelengths that the human eye can detect, so the driver won't even notice it and won't be distracted.
By monitoring eye movements and measuring the distance between recognisable key points on the face, such as the pupils of the eyes and the tips of the nose and ears, a small onboard computer measures the change in those distances and, more important, the rate of change, to gauge how effectively the driver is concentrating on their driving.
This is not theoretical pie in the sky - the technology is already installed in test vehicles, and Landfors' team is co-operating with researchers Chalmers University of Technology in Göteborg to identify effective methods of detecting tiredness and distraction.
The possibilities are endless, says Landfors. Monitoring eye movement tells the car where the driver is looking; from there it's but a small step to having the car's ambient lighting come on only where the driver's eyes are focused.
For instance, the dashboard lights could go out when they're looking in the mirror, and the backlighting on the air-conditioning controls could go on only when the driver is looking at them.
Landfors was quick to point out, however, that the system does not save any pictures or record of movements. The legal ramifications in a country such as Sweden, where civil rights violations are taken very seriously indeed, would be too big to handle.
"Nor does it have a driver surveillance function," he insisted.
"GOOD MORNING, MICHAEL"
But, by extending the facial recognition software to a database of authorised drivers, it would be possible for the car to recognise the person behind the wheel, adjust the seats and controls to their preference, set the climate control to their liking, tune the sound system to their favourite station and wish them a cheerful "Good morning, Michael."
But Landfors foresees a far more serious application; driver-state estimation will be key to the self-driving cars of the future.
"The car will need to be able to determine for itself," he explained, "whether the driver is capable of taking control when the conditions for driving autonomously are no longer present.
"This means the driver will be able to rely a bit more on their car, and know that it will help them when needed."