Mitsubishi Electric uses machine-learning tech to detect distracted drivers

Distracted driving causes a huge number of accidents each year with some of those accidents resulting in the death of the driver, passenger, or pedestrians. Mitsubishi Electric has announced that it has developed machine-learning technology that is designed to help the automotive industry detect distractions in drivers when the vehicles are traveling straight.

The company uses a type of machine-learning algorithm known as deep learning and believes that this is a first for the automotive industry. Mitsubishi Electric says that systems are available that can detect drivers who are drowsy or inattentive, but detecting cognitive distractions is difficult.

The difficulty comes in that symptoms appear in the driver's behavior or biological patterns rather than in face or eye movements. The tech Mitsubishi has developed uses a machine-learning algorithm to analyze data including vehicle steering, driver heart rate and facial orientation to warn drivers of potentially dangerous indicators.

The deep learning tech uses forward and backward moving time-series data to help reduce errors in predicting driver actions by 66% compared to a conventional algorithm tested. Mitsubishi will show off the tech during the Tokyo Motor Show and expects the tech to be installed in driver sensing units sold commercially in 2019 or later.

SOURCE: Mitsubishi Electric