Nauto has added a predictive collision alert module to its driver behavior learning platform that is designed to detect imminent collisions.
The new module helps reduce rear-end accidents by up to 400% more than traditional approaches, according to Nauto. It simultaneously fuses driver behavior, vehicle movement, traffic elements, and contextual data to help predict and prevent collisions
Nauto Predictive Collision Alerts continuously synthesizes inputs from in and around the vehicle, including driver behavior, vehicle movement, traffic elements, and contextual data, in its multi-tasked Convolutional Neural Networks (CNN) model to determine levels of collision risk. As the detected risk intensifies, Predictive Collision Alerts signals the driver to take action with increasing levels of urgent alerts.
The initial implementation of Predictive Collision Alerts will focus on reducing rear-end incidents—which account for an average of 26% of total fleet losses at $82,000 per incident, according to the company.
“With more than 650 million AI-processed miles informing our platform, Nauto already leads the way in providing real-time training that helps drivers adjust behavior and avoid collisions in the moment. Traveling at 60 miles per hour, our new Predictive Collision Alerts could give drivers as much as 100 extra feet to react to a potential collision,” said Shweta Shrivastava, chief product officer, Nauto.
Nauto’s Predictive Collision Alerts will be available for order this June.
Originally posted on Automotive Fleet