Local news recently reported a 28-year-old mother who was killed by a drunk driver going the wrong way on US 60. A sad, but common story about the impact of impaired driving.
Federal regulators are seeking a solution. In January, they published an Advanced Notice of Proposed Rulemaking.
The objective is to solicit information on technology for requiring vehicles in the future be equipped with an impaired driver detection system. In other words, a computer system to stop drunks from driving.
However, the regulators explicitly stated the system must be passive. That is, the driver should not have to do anything “extra” before turning on the ignition, such as using a breathalyzer to start the vehicle.
This requirement comes directly from the 2021 Infrastructure Investment and Jobs Act.
But passivity is proving to be a major hurdle.
In search of a technology
NHTSA is seeking public feedback on this proposed requirement so they can draft suitable guidelines for a new motor vehicle standard. (See Regulations.gov/docket/NHTSA-2022-0079.)
For instance, should the technology be tuned to detect levels above the federal legal limit (0.08 grams per deciliter) or should it be adjustable for states with stricter levels? Could erratic or drowsy behavior be used to identify an “impaired” driver?
While considering a new standard, NHTSA must also demonstrate technologies exist that could meet this requirement. NHTSA is funding related research.
Investigators at the University of Iowa studied in-vehicle driver monitoring systems. The idea is to point cameras, capable of detecting alcohol-impairment, at the driver.
They performed extensive tests with live participants to detect drowsiness, often associated with drunk driving. But no technology rose to the level needed to satisfy a national standard.
Researchers at Battelle Memorial Institute studied 331 passive technologies, including physiology-based (heart rate, respiration rate), touch-based (measuring alcohol through the skin), and camera-based systems (monitoring speed and steering inputs).
In addition, the researchers studied whether automation systems could take over driving once an impaired driver was detected. None of the commercially available systems they investigated qualified.
Asking for the impossible – now
Authorities have used enforcement, public education, and social pressure to combat drunk driving for 50 years. Initially, those programs appeared to work.
Between 1980 and 2010, the ratio of on-the-road fatalities associated with alcohol impairment decreased from nearly 1 in 2 to about 1 in 3. But in the past decade, an average of almost 10,500 people died each year in these driving crashes.
Hundreds of individuals have submitted their personal stories to NHTSA about the far-reaching impact of drunk drivers. Over and over again, the words preventable, tragedy, and murder are woven into the narratives about the tragic loss of children, spouses, and grandparents.
One grieving family member wrote: “Please pass this [install the technology] before YOU become the victim.”
Why can’t carmakers create a driver-assist technology that eliminates impaired driving? If they can create emergency braking to stop, or park a car on its own, why not stop drunk drivers? Again, the passivity requirement makes it complicated.
The system needs to be fine-tuned to people’s individual characteristics and general enough for any driver. Some eyes look sleepy. Some people jerk the wheel erratically. Perhaps this signals a drunk driver. Or is it just a quirky sober driver?
And once the system decides the driver is drunk, should the car stop in traffic? Should the car self-drive and find a safe place to park?
I am fully in favor of removing drunk drivers from the road. But our society, and our lawmakers in particular, are once again asking technology to do too much.
To fix a social problem that humankind finds unbeatable. Is a new computer system really the answer? I am unconvinced. For now.