Believe it or not, fleet safety managers are sometimes wrong. With a goal of reducing risk in their operations, more fleets are adopting video-based recording systems to identify risky driving behaviors. Things like hard braking and lane departures are indications of unsafe driving.
A safety manager sees a report come across his screen that Driver X just had a hard-braking event and he springs into action. An email or text may be sent, a conversation had, even a possible suspension. But what if that driver did the correct thing under the circumstances? The video doesn’t lie, but it might not tell the whole story.
That’s where context can play a pivotal role, and to get that context, you need more information. Netradyne is using artificial intelligence (AI) to provide fleets with the context needed to make more informed decisions.
“The technology behind what we are doing [is different] compared to what has been on the market,” Adam Kahn, vice president of fleet business for Netradyne, explains. “Instead of a legacy system that turns on when an incident happens, it scans the roadway continuously.”
Netradyne’s system is “vision-based” and not only monitors the roadway with three cameras, but also can read road signs and identify stoplights.
“We can start to make very accurate assessments of what happens on the road and do that at the device level,” Kahn says. “It changes the relationship between the fleet manager and driver.”
By using AI, the Netradyne system can provide context to an event. For example, Kahn mentions a real-world example of a Netradyne client whose driver had a hard-braking event. Using a traditional event-triggered system, that driver would have a hard-braking event on his log. The Netradyne system, though, logged the incident as a hard-braking event, but because of the AI, it was able to add context to the situation, in this case there was a car stopped in a lane adjacent to the one the tractor-trailer was in and a woman opened the door on the car into the driver’s lane. An unexpected event that was recorded as hard-braking due to a third-party incident.
Each system sends a notification to the fleet, but in the case of the Netradyne report, the fleet manager knows, because of the context, that his driver reacted appropriately to the situation. The fleet could then choose to reward drivers who are reacting positively to avoid incidents rather than just penalizing drivers who have such triggers.
“Now we’re giving the safety manager the information to reward drivers for doing the right thing,” Kahn notes.
The notification system is part of Netradyne’s DriverStar program, which was just released. The entire system utilizes up to four high-definition cameras – a front view, two side cameras that caption of the A pillars, and an optional driver-facing camera. In total, they provide a 290-deg. view of the vehicle, Kahn says.
“I might be tracking dynamic following distance, but the system is also logging [vehicles and items in adjacent lanes],” Kahn adds.
A high-powered GPU with a teraflop processor capable of making a trillion calculations per second is included and the system utilizes “machine-learning” so it can identify items over and over again.
DriverStar alerts fleet safety managers of identified great driving decisions made by their commercial drivers immediately after they happen, providing safety managers the power to recognize and reward extraordinary driving behavior.
DriverStar is an extended feature to GreenZone, which already tracks driving behavior throughout the day in the Driveri system—constantly analyzing and assessing performance (e.g. compliant driving, traffic light compliance, following distance, stopping at stop signs, etc.).
Its AI-capabilities provide a deeper view into positive driving behavior, analyzing data and alerting safety managers when a driver reacted safely to a challenging situation—reducing the risk of a crash or collision.
Stay up-to-date with the latest commentary and insights on FreightTech and the impact to the markets by subscribing.
