Are you okay to drive? Your eyes may answer

These days, cars know a lot about the behaviors of their drivers.

They know when motorists are traveling faster than posted speed limits, and they know when drivers stray from their lanes. With the advent of driver-monitoring systems, they know when drivers glance downward and when their eyes wander from the road ahead.

This influx of information can help mitigate human errors and improve automotive safety. But these methods fail to address a more fundamental component of the safety equation.

"Everybody talks about safe driving," says Carl Pickering, CEO of ADAM CogTec. "But nobody asks the question: 'Is the driver safe to drive?' "

To date, the only answer might come from an ignition interlock device that detects alcohol impairment. But Pickering and others at the Israeli startup have been developing technology that can unearth a trove of information by peering into a driver's eyes.

With a driver-facing camera, the company's product can analyze a driver's cognitive condition through ocular responses to a series of flashing lights, determining whether the person is impaired by drugs or alcohol, just plain fatigued or some combination of all three.

The software augments current systems that monitor head position and track eye gaze — indeed, ADAM CogTec's path to market includes partnering with suppliers of these traditional driver-monitoring systems.

Those relationships are forming at an important time. By 2025, 8 percent of new vehicles sold in the U.S. will have a Level 2 driver-assist feature, and of those, 100 percent will contain a driver-monitoring system, according to projections from global automotive consulting firm SBD Automotive.

Enhancing current driver-monitoring systems with more sophisticated software could be a compelling safety proposition, says Alain Dunoyer, head of autonomous car research at SBD.

"There are cases where you could be looking straight ahead, but your attention is not where it should be, and you can be daydreaming," he said. "That's where the current technology is not sufficient. ... [ADAM CogTec seems] to be able to potentially evaluate the mental workload of the driver, which is really the end game."

Internal tests show ADAM CogTec's software can improve driver attention and response times as much as 60 percent, according to the company. ADAM CogTec, founded in 2017, received $2 million in funding from venture capital firm Mobilion Ventures in May. The fund is backed by global companies such as Nexus Automotive International, Delphi Technologies and Mitsubishi Corp.

Pickering said the company will have prototypes installed as aftermarket products in commercial fleets by the fourth quarter. Testing and validation will continue through next year, with a tentative plan to launch the system either toward the end of 2021 or in early 2022.

ADAM CogTec's software establishes baseline results for individual drivers, then can detect deviations from those normal indications when a motorist is impaired or fatigued.

Frequency of testing can be tailored by car companies or fleet operators. Some might check a driver's cognitive state when the person enters the vehicle. Others might at regular intervals or initiate a scan after an unusual event, such as a hard braking.

More novel use cases may exist. Imagine climbing into the back of an Uber or Lyft, for example, and seeing the driver's current test compared with their normal baselines. Applications stem beyond the auto industry into other transportation sectors and even the health industry.

"Take a surgeon who's about to enter a five- or six-hour triple-bypass operation," Pickering said. "Where's the check that says he's safe to operate? Imagine on the entrance door to the operating theater there's a real-time, noncontact system that gives him the green light or an amber one that says, 'You didn't sleep much last night.' Anything that indicates a decline of cognitive performance against the norm."

Before joining the startup, Pickering was Jaguar Land Rover's head of autonomous technology strategy and global human-machine interface manager. During his tenure, JLR experimented with measuring cognitive state by using thermographic images of a driver's head to monitor blood flow. Researchers learned blood flow increases in a variety of scenarios, including when a driver gets embarrassed.

Throughout the industry, there has been research that attempts to infer a driver's emotional state and cognitive condition by analyzing facial expressions.

"But that can only go so far," Pickering said. "Some of them can monitor when the eyes are starting to close, and they can pop a drowsiness warning. But that's as close as they can get. So this layer on top is a natural fit."