Companies developing driver detection technology could get a boost from a provision tucked inside the 2,702-page $1 trillion infrastructure bill that would require automakers to build into new cars technology that can tell if drivers have had a few cold ones.
The provision in the bill, which is actually a piece of bipartisan legislation called the Reduce Impaired Driving for Everyone Act that was introduced in April 2021, would direct the U.S. Department of Transportation to establish a technology safety standard for automakers within three years. Automakers would then have another two years to comply and implement tech that detects and prevents drunk driving. Reuters was the first to notice the language.
While the provision doesn’t dictate what type of tech has to be in these vehicles, industry experts believe that companies developing camera-based driver monitoring systems (DMS) stand to benefit the most. DMS systems are already mature in the auto industry, representing a technological byproduct of autonomous driving developments. While the auto industry explores self-driving cars as a way to drastically reduce road deaths in the future, advocates and regulators say there’s room to use some of this tech to solve problems that exist now, like drunk or distracted driving.
“What’s happening in the U.S. Senate this week potentially opens the door to a camera-based real-time solution, which will be the first time that the U.S. automakers will have the ability and the requirement to look at real-time physiological changes in your body that occur when you are inebriated,” Dr. Mike Lenné, chief science and innovation officer at Seeing Machines told TechCrunch. “There are distinct reliable changes to the way you scan the environment, to the way your eyes respond to stimuli, which is why the police use that ‘follow the finger’ test.”
The system would have to monitor the performance of a driver to detect impairment and prevent or limit vehicle operation if impairment is detected; detect whether BAC (blood alcohol concentration) is equal to or greater than the legal limit, potentially preventing operation of the vehicle at all; or a combination of both systems.
Cameras aren’t the only solution that has been trotted out in recent years.
The Driver Alcohol Detection System for Safety (DADSS) program, a technology that’s been developed in partnership between the Automotive Coalition for Traffic Safety and the National Highway Traffic Safety Administration (NHTSA), has advocated using a breath or touch-based approach to determine BAC levels. The touch-based approach involves measuring BAC through the skin’s surface by shining an infrared light through the driver’s fingertip. According to DADSS, the current timeline for bringing the breath-based approach to vehicles is by 2024, and the touch-based approach by 2025.
Lenné argues that a camera-based approach would be far more successful than a breath or touch-based approach because BAC levels can rise within minutes. Someone could theoretically down a bunch of shots immediately before getting behind the wheel and it wouldn’t show up on a reading for several minutes. Or they could get wasted while driving. And BAC detection doesn’t help at all when it comes to drug-impaired driving.
Europe versus U.S.
Moves are already being made in Europe to encourage automakers to include drunk driving detection technology, specifically through camera-based DMS approaches, whereas most of the discussion on this type of tech in the U.S. has been, until recently, focused on DMS for assisted driving and Level 2 autonomous driving and above. (According to the Society of Automotive Engineers, Level 2 autonomy means the vehicle has combined functions like steering and acceleration but requires the driver to remain engaged.)
The U.S. provision could propel an industry that has already seen growth in recent years as automakers like GM and Ford implement hands-free advanced driver assistance systems.
“From an integration viewpoint, it’s actually not a step change at all from what the OEMs are doing right now for distracted driving and drowsy driving with camera-based DMS. It’s just another feature to offer, another algorithm on the chip, if you like,” Lenné said.
“Billions of dollars have gone into developing the technology to make AVs a reality but they are really far off,” Stephanie Manning, chief government affairs officer at Mothers Against Drunk Driving (MADD), told TechCrunch. “In the process, automakers have developed a lot of technology that can help us right now in terms of saving lives. If this passes, it’s going to be the biggest safety rulemaking that NHTSA has ever done in terms of lives saved, and it couldn’t have come at a better time. But the more we wait, the more we delay, the more people die.”
The technology is not at all far from market, said Lenné, and he would know. Seeing Machines provides the DMS that is used in Super Cruise, GM’s hands-free advanced driver assistance system. Super Cruise, once relegated to just one Cadillac model, has expanded in capability and GM’s portfolio and is now in the Cadillac CT6, CT4, CT5, Escalade and Chevrolet Bolt. Seeing Machine’s tech is also used in the new Mercedes-Benz S-Class and EQS sedans.
“Once it’s regulated, we can expect to see more entrants to the market because what this does is it creates a top-down demand,” said Lenné. “It takes it out of the consumers’ hands and tells vehicles they must have these safety features, so the market size will increase dramatically, and so will the market opportunity.”
The global DMS market is estimated to surpass $2.1 billion by 2026, growing at a compound annual growth rate of 9.8% from this year, according to IndustryARC. Top-down demand due to regulations like the infrastructure bill will certainly increase demand, but it won’t make the problem easier to solve.
“We’re trying to assess what’s going on in someone’s head, and that’s really different from having a forward-facing radar that’s trying to look at what’s 30 meters in front of you,” he said. “You’re trying to interpret whether or not this person is safe to drive. So it’s a really difficult technical problem to solve. Our company is 21 years old. Smart Eye has been around for over 10 years. Whilst the market size has increased dramatically, it’s a hard problem to solve as a new entrant.”
Newcomers will face competition from established and large Tier 1 suppliers like Seeing Machines and Smart Eye, a Swedish computer vision company that people familiar with the industry say works with Ford (Ford did not confirm or deny this). IndustryARC also names major players as Faurecia, Aptiv PLC, Bosch, Denso, Continental AG and others. But new players are finding their way into the scene, like Israel-based Cipia, formerly Eyesight Technology, and Sweden-based Tobii Tech.
Room for growth in the market
More entrants to market means more advancements to the technology. Smart Eye’s recent acquisition of emotion-detection startup Affectiva for $73.5 million hints at the potential future applications of DMS in passenger vehicles. Today it might be distracted, drowsy or drunk driving, but in a few years DMS could detect other types of drug impairment, cognitive impairments or even road rage.
Tobii, an eye-tracker technology company, just announced its entrance into the DMS market, a space it’s been exploring for the past few years as it has watched the legislative changes happening first in Europe and now in the U.S.
While a new entrant to the automotive space, Tobii has been in the eye tracking space since 2001, working in industries like marketing, scientific research, virtual reality, gaming and more. Anand Srivatsa, Tobii’s division CEO, told TechCrunch he thinks one of the biggest challenges will be scaling across different populations, given the different eye shapes of different ethnicities, which he says puts Tobii at an advantage, even with its limited automotive experience.
“Because of this long history, we have what it takes to deliver a full solution from a component level all the way to end software because we’ve done it in other parts of our business,” Srivatsa told TechCrunch. “Some of our automotive partners see that as a unique capability from Tobii where we can talk about the compute that is needed for eye tracking because we build our own asix, we’ve built our own sensor. We have end user software in some aspects of our business, so we understand the implications and the constraints of each of these parts of the stack, and we can work with them to create a more disruptive solution. And that’s something that I think is going to be quite important in this space. How do you reduce the total cost of the solution to allow it to scale efficiently across all cars?”
Srivatsa also said there’s room to extend into other spaces the biometrics or physiological signals that eye tracking yields, reconfiguring information based on outside road conditions or what else is going on in the car in a way that optimizes the tech to ensure drivers are spending the bulk of their time looking at the road.
“What I am hoping and dreaming for is technologies like forward collision warning, or blind spot warning or even the lane swerving warnings help me out when I need it most by understanding if I’m becoming complacent or tired, perhaps distracted, and then adjust how the systems perform, the warning timing and things like that, based on what I need in the moment,” Kelly Funkhouser, program manager of vehicle interface testing and head of connected and automated vehicles at Consumer Reports, told TechCrunch. “Counter to that is I would like it to not bother me and nag and annoy me when I am fully paying attention. I’m like ‘Yeah I know exactly what I’m doing, I am purposely driving over this line so that I don’t hit the mom and kids.’ ”
Lenné said there’s a potential for driver monitoring systems that capture what is really going on inside of a car to become more personalized in order to provide a better driving experience.
“I think in all of this, writing a better driving experience is absolutely pivotal,” said Lenné. “If it doesn’t do that, it risks not being accepted by the consumers.”
Advancing existing ADAS tech
Automakers have been a part of the conversation regarding drunk driving technology for years. Back in 2007, Nissan revealed a drunk driving concept car that would use alcohol odor sensors, facial monitoring and vehicle operational behavior to detect driver impairment.
In the same year, Toyota announced a similar system that it said would be in cars by 2009. More recently, Volvo announced in 2019 that it would install cameras and sensors in cars to monitor drivers for signs of being drunk or distracted and then signal the vehicle to intervene, but that tech is designed for Volvo’s SPA2 architecture for hands-free driving, which hasn’t been released yet. The bottom line is without legislation mandating drunk driving prevention and detection, automakers haven’t really moved forward on implementing the tech, despite much of the building blocks being in place already.
Manning thinks that’s because automakers want to be able to upcharge for safety features.
“Automakers want to test their supercomputers on the open road, but they don’t want to put the money and time and energy into solving drunk driving, because they don’t feel it’s their responsibility, and they don’t want this rule-making,” she said. “We fully expect that they’re going to fight us tooth and nail throughout the rule-making process.”
Representatives from GM and Ford could not be reached for comment, but John Bozzella, president and CEO of the Alliance for Automotive Innovation, which worked with NHTSA on the DADSS program, told TechCrunch that the auto industry is committed to supporting public and private efforts to address this threat to road safety.
“We appreciate the efforts of congressional leaders and other stakeholders to advance a legislative approach that provides NHTSA the ability to review all potential technologies as options for federal regulation and, consistent with the Motor Vehicle Safety Act, to make a well-informed decision as to whether any specific technologies meet the standard for consumer vehicles,” he said.