PALAKKAD: What if a steering wheel could sense a driver’s hands without seeing them? Researchers at IIT Palakkad have developed a novel touch-based sensor system that can detect where and how a driver holds the wheel, without using cameras or capturing images.
Opening up new possibilities for safer and more privacy-conscious human-machine interfaces, the technology, the developers said, could pave the way for safer vehicles while addressing privacy concerns associated with camera-based driver-monitoring systems.
Developed by Sreenath Vijayakumar, an associate professor, and Pinku Sebastian, a technical superintendent, of the department of electrical engineering, the system uses specially designed sensor strips integrated into surfaces such as steering wheels.
By detecting the electrical signals generated by human touch, it can identify hand positions and distinguish between two-handed, one-handed, and prolonged no-hand driving, they said.
The technology can also detect variations in grip.
“The interface can instantly detect where exactly your hands are placed around the wheel and how firmly or lightly you are holding it. Such information could be used to provide warnings in situations involving distraction or potentially unsafe driving behaviour,” Sreenath said.
Unlike camera-based driver-monitoring systems, the sensor does not capture images of the driver.
This eliminates concerns associated with visual surveillance while also allowing the system to function irrespective of lighting conditions. It can operate in darkness as well as bright sunlight and is less affected by situations in which a driver’s hands are temporarily obscured from a camera.
“While the cameras track the driver’s movements, our technology is based on the touch signals. It can sense whether a driver is losing grip on the steering due to fatigue, sleep, carelessness, etc. There is a set of standard reading in the system, and once that changes, the system automatically responds to it,” Pinku said.
The researchers tested the technology by developing a prototype steering wheel fitted with the custom-designed sensors.
Tests demonstrated the system’s ability to identify hand positions with high precision, while its relatively simple construction makes it considerably lighter and potentially less expensive than camera-based monitoring systems.
The applications of the technology could extend well beyond automobiles.
According to the researchers, the same touch-sensing principle could be incorporated into helper robots to enable them to detect human contact and respond appropriately, potentially reducing the risk of accidental collisions.
The technology could also find applications in healthcare. Its integration into smart ICU beds, for instance, could help monitor patient movements and changes in weight distribution and provide alerts when a patient is at risk of falling.