U-M engineers create tech to spot drunk drivers in cars.

TLDR:

• Drunk driving is a leading cause of death on U.S. roads with 13,384 deaths in 2021.
• U-M engineers have developed technology using cameras and LiDAR to detect impaired drivers.

According to the National Highway Traffic Safety Association, drunk driving continues to be one of the leading causes of death on U.S. roads, with 13,384 deaths in 2021. In an effort to reduce these statistics, engineers at the University of Michigan (U-M) have created technology that can detect impaired drivers. The technology uses Advanced Driver Assistance Systems (ADAS) cameras already present in newer vehicles, combined with infrared LiDAR or structured light 3D cameras, to monitor drivers for signs of impairment.

The U-M team’s proof of concept experiments involve detecting five markers of impairment, such as blood flow, eye behavior, and respiratory rate, using artificial intelligence tools. The system works by analyzing data from the cameras to identify anomalies that may suggest a driver should not be driving, similar to how a credit card company detects fraudulent transactions. The goal is to get this technology out of the university and onto the roads with the help of local support and financial backing.