The Road to Solving Impaired Driving Using Artificial Intelligence

Impaired driving is a phenomenon that encompasses a variety of instances – alcohol intoxication, drugs or sleep deprivation. Acusensus is developing AI-based technology to combat impaired driving. A collaborative project with the Federal Office of Road Safety and Griffith University, this solution will be able to measure a driver’s attentiveness, reaction time and control in real-time to determine impairment levels. This will assist the random process of random breath tests and help narrow in on behaviours with more accuracy.

In the last five years, approximately 41% of all drivers and motorcyclists killed who were tested had drugs in their system and 20% had a BAC over the legal limit of 0.05. The Acusensus Impaired Project aims to reduce fatalities attributed to impaired driving using external AI systems set up in fixed cameras or trailers. This project identifies a research gap in the world of road safety as there are few attempts to create an external fixed solution for the detection of impaired driving.

The system has been trained to detect features synonymous with impaired driving. If a driver was detected showing signs of being affected by drugs or alcohol, police would receive a notification ping. Using that information, police could then direct the driver to a breath test or test the person for alcohol or drugs.

Acusensus plans to proceed to a real-world trial stage following successful simulator testing - this process will be conducted alongside the Queensland Police Service, making the state the first to adopt the technology. Acusensus hopes to extend this technology wherever ‘Heads-Up’ solutions – mobile, speeding and seatbelt AI systems – are available in Australia, which is NSW, Queensland, Western Australia and the ACT and hopes to eventually deploy the technology worldwide.

Video