DIME – Using smart technology to stop impaired driving and save lives
How can we detect if a driver is under the influence – even when they try to hide it? That is the challenge at the heart of the new partner project Driver Impairment Multimodal Evaluation (DIME).
Led by the University of Gothenburg together with Smart Eye, DIME aims to develop next-generation technology for recognizing signs of impairment in drivers. The goal is clear: to reduce alcohol-related crashes and bring us one step closer to Vision Zero – a future with zero fatalities and serious injuries in traffic.
Saving lives through smarter technology
Funded by VINNOVA (FFI), the project runs from August 2025 to August 2027.
Using deep learning and cutting-edge data fusion, DIME combines three key types of information:
- In-cabin driver behavior – how the driver moves and reacts,
- Speech patterns – voice and language cues linked to intoxication,
- Biobehavioral signals – such as heart rate, eye movement, and steering patterns.
Together, these elements will create a powerful multimodal system capable of identifying intoxication even when a driver actively tries to conceal it. The system is planned for commercial deployment by 2028.
Collaboration through SAFER – turning research into real impact
Within the SAFER community, DIME offers a strong opportunity to create added value through collaboration. By connecting experts in vehicle safety, human factors, and AI, SAFER will help ensure that research outcomes lead to real-world applications that improve safety on the road.
Throughout the project, knowledge-sharing workshops and community discussions will be organized to spread insights, encourage cooperation, and inspire new innovation based on DIME’s findings.
“We want to learn together, share insights, and make sure that research results are translated into real safety benefits,” says Robert Lowe, Project Lead at University of Gothenburg.
