Imagine if automotive regulation allowed seat belts to unlatch or airbags to fail to deploy in over half of severe collisions. And legislators were dragging their feet on raising the minimum standard for fault tolerance. In such a scenario, who should fix the problem?
Thankfully seatbelts and airbags are mature, well understood technologies, and the laws mandating them are sufficiently stringent. For newer safety technologies, however, this is not the case, and legislators aren’t moving as quickly as they should. The more I think about it, the more I believe that the smart original equipment manufacturers (OEMs) will drive the change required to save more lives on our roads. They will achieve this by being selective in which technologies they put in the vehicle and subjecting the claims of tier 1s to more scrutiny.
I’ve been active in the automotive sector for years, and more recently in the in-cabin space. I recently returned from the InCabin Brussels trade show. The safety innovations in-cabin comprise cameras and other sensors that monitor conditions like driver drowsiness, intoxication, and distraction, as well as child presence detection or seat belt recognition. These are all vital technologies that will save lives. But I see obstacles to them reaching their potential.
What has stunned me is how many mediocre safety technologies are finding their way into vehicles. As we walk the trade show floor, we chuckle at the various errors these systems make such as classifying a fully alert yawning person as drowsy or waiting until a driver is slumping and clearly incapacitated before triggering a warning. But on reflection, in the context in which they are required, it feels a bit macabre to laugh. I have had a long think about what is wrong, and how we can fix it. This would help us achieve Vision Zero of eliminating all road fatalities and severe injuries on our roads.
Why don’t some OEMs select the best technologies?
Regulation is the main driver used to push new safety technologies to market. It sets a minimum standard for compliance. Problems arise when the bar is set too low. I’ll discuss drowsiness because I am most familiar with it, but this problem exists across many emerging safety technologies. In the case of drowsiness detection, the EU General Safety Regulation mandates a mere 40% sensitivity. This means that a system can miss 60% of the times the driver is dangerously drowsy and on the verge of having microsleeps. Once again, imagine if in over half of severe collisions seat belts were allowed to unlatch or airbags could fail to deploy!
From an automaker’s perspective, safety technology merely needs to be compliant. Anything that is vastly safer than the bare minimum offers little short-term advantage. It still results in the same ticked box. And in most cases, the technology in question is too complex to communicate a point of differentiation to end consumers. This is especially the case in driver monitoring systems, which focus specifically on mitigating driver error and could be seen as patronising. Consequently, most auto brand messages are centred on upbeat lifestyle concepts, rather than safety features. Gone are the days of ads with crash test dummies.
If an OEM merely chases short-term KPIs, spare resources will usually be focussed on comfort and entertainment features because they are better understood by end consumers. Consequently, some automakers are not selecting the safest technologies because the legislation is too lenient. A side effect is that consumers will erroneously think they have a system that protects them against dangerous situations with high accuracy, rather than less than half the time.
|“I know I feel dangerously tired, but the car will alert me when it’s a real risk. Surely they wouldn’t be allowed to sell me something that is less accurate than a coin toss.”|
Why don’t tier 1s always choose the best solution?
My team at Optalert offers technology that quantifies impairment from drowsiness. Initially I thought my job as a marketer was to communicate our technological superiority to the industry. If automakers knew we had the best technology, they would put us in their cars, right? What I’ve since learnt is that safety technology that enters the vehicle merely needs to be compliant. Anything that is vastly safer than the bare minimum offers no short-term advantage. They both result in a ticked box.
We regularly speak to tier 1s who supply the major OEMs. They repeatedly tell us they know our system is objectively superior. So there we go: I’ve done my job. They’ve seen the evidence and know we’re the best. Now they should queue up to integrate us into their tech stack. We don’t increase cost. It’s a no-brainer, right?
But most of them have achieved compliance and are resource-constrained. Cost is not the limiting factor. It’s usually a combination of the following:
- Not many OEMs are asking for it
- Reluctance to add another supplier to their technology stack
- Sunk cost bias from having invested so much in an internal drowsiness project
- Lack of understanding across the business of how to objectively compare systems
- Internal politics from leaders defending their fiefdom over selecting the best technology
- No time to swap in better technology because they are frantically scrambling to deliver the current backlog of orders
Figure 1. The dilemma that emerges when regulators set too low a bar for automotive safety technology
How will this collective action problem be resolved?
|How many road fatalities would we prevent each year if we mandated a minimum of 80% sensitivity in drowsiness detection?
Around 20,600 people were killed in road crashes in Europe last year. Drowsiness and sleep have been found to cause somewhere between 16% and 24% of vehicle accidents. Further, they result in higher rates of fatalities because the driver doesn’t slow down or take evasive action. But again, let’s be conservative and settle on 20%, which equates to 4,120 road fatalities per year in Europe due to drowsy driving. In a scenario in which all vehicles prevented drowsiness-related accidents 40% of the time, this would save 1,648 lives per year. If the legislative minimum was increased to 80%, it would save an extra 1,648 lives per year.
The technology exists now to achieve a sensitivity well above 90%, but to be conservative let’s advocate for a minimum of 80% sensitivity and consider its impact against the current minimum of 40%. According to my back of the envelope calculations (see breakout box above), this would save an extra 1,648 lives per year in Europe.
In simple terms, the current law is better than nothing. It will reduce European road fatalities by about 8%. But we can double its impact sooner. No new technology is required – it exists today.
The smart OEMs are pushing for this better safety technology now, which will deliver clear advantages in the medium term:
First, they avoid the public backlash from ineffective safety technology. There will inevitably be scrutiny on DMSs that failed to respond to an obvious case of driver distraction or impairment in the near future. Don’t be that OEM. Instead, acquire their customers.
Second, their customer experience will not be marred by annoying false alerts from an erroneous safety system. Keep an eye out for tier 1s that sound quiet alerts – this indicates that they are not confident in their drowsiness or impairment algorithm. If a system is truly confident the driver is dangerously impaired, the alarm should be loud enough to wake them from sleep or near sleep.
I am confident the negligible short-term cost savings from taking tier 1 claims at face value will be negligible compared to the longer-term loss of market share from integrating weaker technologies. The collective action problem will not be resolved through more stringent legislation, but rather by weaker technologies being selected out of the market.
This may take longer than we would like. But in the long run, the smart OEMs will get ahead by toughening their KPIs. And farsighted tier 1s are already anticipating this inevitable shift and moving beyond mere box-ticking. If you are one of the prudent ones, hold firm. Don’t let the short-term successes of corner-cutters dissuade you. You will win in the long run.
Let’s work together
If you’re in the automotive industry, visit our Automotive page and learn about how you can integrate our drowsiness measurement technology via a software development kit (SDK) into your driver monitoring system (DMS). Together we can improve safety on the roads and move toward Vision Zero.