a skilled, alert driver can adapt to a range of traffic conditions – sudden stops in front, swerving vehicles, icy roads, massive potholes. Drowsy or distracted drivers? Not so much.
Telling the difference between the two is going to be critical as automakers begin designing advanced safety systems and autonomous-driving technology that shift some maneuvering tasks from the driver to vehicle systems. Frank Morese, CEO and chief technical officer of Nevada’s Olea Sensor Networks, thinks he has the technology to identify which drivers need assistance and which ones are fully engaged in driving – a heart-rate monitor embedded in car seatbelts.
“Monitoring physiological arousal information of a human being, it’s a very complex set of measurements that tend to depend on heart rate variability. You can look at heart rate and respiration rate as well,” says Morese. “We have been doing a lot of research and collaborating with the auto industry to understand stress. One of the biggest applications is in autonomous vehicles. They want to make sure that people are alert and understand the human-machine interface.”
Olea’s OS-3005 system is being offered as an aftermarket product, but Morese says several manufacturers are considering the technology to power future advanced safety and autonomous driving systems. The sensor uses micro-Doppler radar to measure drivers’ heart beats and respiration. Sophisticated software evaluates heart rate and respiration patterns to monitor a driver’s condition across time.
More often used in clinical settings, the micro-radar systems require powerful software for vehicle use because of difficult monitoring conditions. Car cabins are vibrating, noisy places, prone to sudden jolts as drivers move over rough pavement.
“Any kind of sensor is prone to motion artifacts from the automobile. We have spent a bit of time working around that,” Morese says. “The processor technology and the advanced analytics available today with sensors allowed us to do some things that even five years ago, I didn’t think possible.”
The key, he adds, is that the system knows what it’s looking for. Heartbeats are regular events, and the sensors know when and where they should be audible. Rather than trying to find a random signal in a noisy environment, the system effectively confirms that the regular signal it’s expecting is present. As heart rates change, software adapts and registers the change in physiological driver states.
The biggest opportunity for advanced biometric monitoring, Morese says, is determining driver alertness. Several automotive suppliers have shown systems in recent years that measure drowsiness by watching a driver’s eyes, looking for drooping eyelids. Morese says that method is flawed because people lose alertness long before they begin closing their eyes.
“We’re not in the sleep-detection business. We forecast sleep,” Morese says. “By the time you detect changes in rapid eye movement, the driver is already in a sleep state. It’s been shown that it’s very dangerous to interact with the driver at that stage. People will overcompensate sometimes when receiving alarms.”
Slowing respiration and heart rates tend to show sleep states earlier, he adds. If a car understands a driver’s baseline heart and respiration rates, it can predict, as many as 10 minutes before a driver loses alertness, that he’s getting drowsy.
“You have a system that can warn a driver before you reach what we call a meditative state, when the unconscious mind takes over,” Morese says.
Such systems could be particularly attractive to commercial trucking companies and fleet customers such as taxicab companies. Identifying fatigued drivers could allow systems that could automatically tell certain drivers that it’s time for a break or time to call it a night, reducing the risk of accidents and lowering liability.
Growing interest in autonomous driving systems and advanced safety have all major automakers considering biometric systems, Morese says, but for now, it’s something that’s only being studied.
“It can take years to get something like this into a vehicle. General Motors might put it in Cadillac, but it can take years or even decades to get rolled out across all products,” Morese says. Olea is offering the sensor as an aftermarket product to get it out there, and because automakers often choose future systems from that market.
He adds that embedding the sensor in the seatbelt during manufacturing could improve the device – offering direct power connections, for example, instead of using a rechargeable battery.
“They all have interest in providing this in their vehicles. It’s a high priority to monitor vital signs,” Morese says, adding that he hopes to see some 2017 model year vehicles with embedded sensors. “Timing is the question. Taking a risk on new technologies is always a challenge.”
Olea Sensor Networks
About the author: Robert Schoenberger is the editor of TMV and can be reached at 216.393.0271 or firstname.lastname@example.org.