"Why does a driverless car need to use a driver monitoring system (DMS)? Because it was a test-level system."
“Why did a driverless car use DMS? Because it was a test-level system.”
Okay, that won’t win me any prizes for humor. Instead let me pose a serious question: If the Uber which struck and killed Elaine Herzberg had been installed with a state-of-the art driver monitoring system (DMS) to detect for operator distraction and drowsiness, would she still be alive?
Herzberg’s death in Tempe, Arizona, in March 2018 is thus far a one-off event. However, the testing of experimental AV technology in close proximity to non-consenting humans on public roadways continues, with little discussion of the ethics of the practice and no agreement on federal guidelines for safety procedures.
According to figures from the World Health Organization (WHO), human drivers are responsible for about 1.3 million traffic-related deaths globally each year. Yes, humans cause too many traffic fatalities. But we don’t race to “save lives” by testing experimental cancer treatments or Covid-19 vaccines on otherwise healthy, non-consenting subjects. So why the double-standards for driverless technology?
In its exhaustive investigation into the death of Elaine Herzberg, the National Transportation Safety Board (NTSB) wrote:
The probable cause of the crash in Tempe, Arizona, was the failure of the vehicle operator to monitor the driving environment and the operation of the automated driving system because she was visually distracted throughout the trip by her personal cell phone.
Distracted, drowsy and impaired human drivers kill themselves and other road users. However, distraction, drowsiness and impairment of the human safety driver are critical issues for test-level AV systems too. As the NTSB investigation highlighted:
Had the vehicle operator been attentive, she would likely have had sufficient time to detect and react to the crossing pedestrian to avoid the crash or mitigate the impact.
As AV technology continues to be tested, validated and verified on public roads, so driverless cars will require driver (operator) monitoring systems. This is an industry-wide issue, but why are the AV developers so reticent about discussing the subject? Why hasn’t the industry joined together and agreed on a best safety practice? Why three years after the death of Elaine Herzberg are there still no federal guidelines for safety procedures despite NTSB’s investigations and findings?
Even Waymo depends on DMS
Writing in a blog post last October, CEO John Krafcik announced “Waymo is opening its fully driverless service to the general public in Phoenix.” In a further blog post last month, Waymo also announced the expansion of its AV testing program to San Francisco.
Waymo’s desire to showcase its technical prowess and the robustness of its AV technology by opening up its service to the public resulted in video footage filmed inside its vehicles being posted in the public domain. This video from two months ago shows a driverless Waymo operating with a safety driver.
The cylindrical bar with the strobing LED illuminators is the Seeing Machines Guardian Backup-driver Monitoring System.
The clear AV technology leader is using the same company which developed DMS for Cadillac Super Cruise in the 2018 CT6 and the 2021 Escalade. But why doesn’t Waymo discuss this subject openly? What is the DMS adoption at other AV developers testing on public roads? Why doesn’t the California DMV require all AV testing permit holders to adopt DMS?
The answers are the same. No one wishes to draw attention to the inconvenient fact that distraction, drowsiness and impairment can affect humans irrespective of whether they are driving vehicles or supervising test-level vehicle automation. Somewhat paradoxically, the development of driverless cars on public roads looks likely to depend on driver monitoring technology for many years, and possibly decades, to come.
NHTSA finally wakes up to DMS
NTSB has for several years warned about the dangers of distracted driving, the risks of automation complacency resulting from the use of automated driving systems such as Tesla’s Autopilot and the hazards associated with drowsy driving.
There is evidence that NHTSA is finally waking up to the role that robust driver monitoring can play in detecting distraction, drowsiness and impairment in human drivers, starting with its “Request for Information: Impaired Driving Technologies.” Multiple submissions were made, including from Seeing Machines, Veoneer and Mothers Against Drunk Driving (MADD).
Mike Lenné, Chief of Human Factors & Safety at Seeing Machines, wrote:
Given that eye movements are both essential for driving and highly susceptible to states of fatigue, distraction and intoxication, various parameters of gaze and blink metrics are integral signals for the detection of driver impairment.
Humans collectively drive an almost incomprehensible distance, with many estimates of vehicle miles traveled on public roads totaling over 10 trillion per year. The number of passenger cars and commercial vehicles in use globally totals around 1.5 billion, meaning the fastest way to reduce road deaths is the mass deployment of proven, cost-effective, safety technology which can make human drivers into safer drivers.
In the years and decades ahead, DMS will play a critical role improving safety both in human-driven vehicles and in monitoring safety drivers testing AVs on public roads. The DMS market was discussed at the EE Times virtual conference “The Roadmap to Next-Gen EV & AV” in the keynote called “Why Do We Need Driver Monitoring System?” and followed by an in-depth technical panel discussion on driver monitoring technology.
This article was originally published on EE Times.
Colin Barnden is principal analyst at Semicast Research and has over 25 years of experience as an industry analyst. He is considered a world expert on market trends for automotive vision-based driver monitoring systems (DMS). He holds a B.Eng. (Hons) in Electrical & Electronic Engineering from Aston University in England and has covered the automotive electronics market since 1999.