This AI-powered camera detects a driver's state of intoxication

Researchers at Edith Cowan University in Australia have developed a new technology that could transform road safety.

By using a camera paired with artificial intelligence (AI), this system would detect drivers' intoxication levels even before they take to the road. While this development may prevent road accidents, it also raises significant questions about privacy and the accuracy of this technology.


The device, introduced at a conference on computer vision systems, analyzes videos of drivers based on features such as facial traits, gaze direction, and head position.

According to Ensiyeh Keshtkaran, a PhD candidate at Edith Cowan University and the study's lead author, this system has an accuracy rate of 75%, indicating that further improvements are needed before it can be widely deployed. To train their algorithm, the researchers used videos of 60 individuals at varying levels of intoxication, recorded in a simulated driving environment.

Although this technology appears promising, questions remain. First, the protection of data collected by these cameras. Even though the analysis is automated by AI, the risk of personal data leaks is never entirely eliminated. Additionally, such technology must still demonstrate its reliability to avoid incorrectly preventing a sober driver from starting their vehicle.

Despite these challenges, the researchers emphasize that this system could complement existing devices like ignition interlock breathalyzers. These are already used in some vehicles to prevent driving under the influence of alcohol, but they require the driver to interact directly with the device. The system proposed by the Edith Cowan University researchers could be more passively integrated into road surveillance cameras, providing an additional means to prevent accidents.

Finally, the potential use of this system by law enforcement could mark an advance in the fight against drunk driving. However, its widespread adoption will depend on resolving the ethical and technical questions surrounding this technology. To be effective, it must not only be accurate but also respect the individual rights of drivers.