Artificial intelligence can help drivers subconsciously stay away from reckless speeding

Photo: Shutterstock

Photo: Shutterstock

A recent report on highway safety by the Governors Highway Safety Association (GHSA) revealed that speed-related accidents accounted for approximately one-third of total road fatalities. Speeding can result from several different factors including fatigue, alcohol-impaired driving, high-stress levels or simply because the drivers subconsciously believed they were driving within speed limits.

“Speeding remains a publicly accepted driving behavior that is reinforced among motorists, policymakers and transportation stakeholders. National surveys of U.S. drivers have found that although drivers identify speeding as risky, drivers nonetheless continue to speed. Drivers have a minimal perception of risk of either getting a ticket, causing a crash, or violating social norms,” the authors of the report observed.

With regard to the trucking community, though its percentage of speeding occurrences compared to that of passenger car drivers has traditionally been lower, there still is a room for improvement. FreightWaves spoke with Adam Kahn, vice president of fleet business at Netradyne, discussing the potential of technology to avert road accidents due to speeding.

Kahn explained that the speeding traits of truckers can be eliminated with the help of artificial intelligence (AI) and edge computing. Kahn’s company Netradyne uses a vision-based driver recognition tool to help drivers understand their reckless-driving behavior and incrementally get better.

The tool uses both inertial-based and vision-based sensors to gain visibility of the road environment. The data is fed into the tool's AI and edge-computing algorithms that process it to provide feedback in near real-time to drivers. “One of our fleets provided the tool to its drivers, and compared the results from the previous month to the current month. The management saw an immediate 13 percent reduction in speeding – all because we created visibility to the driver on their driving behavior,” said Kahn.

Kahn insisted that the feedback relayed to drivers must be mellow and possibly encouraging for it to yield maximum impact. Psychologically, people are tuned to weed out critical remarks, and thus an appreciative tone on the feedback will cajole drivers to get better at the job. “There’s plenty of research that talks about the power of praise versus the power of punishment. And we’ve seen that if you can base your conversation around the totality of the system and the driver’s performance, the coaching method is very well-received,” he said.

Kahn also spoke about why speeding happens and what can be done to avert it. “We studied the driver community and the risk factors they are associated with and classified them in different ways. At first, we look if the speeding is on the highway or in a residential area, and then look at the context of speeding. For instance, speeding when you are the sole person on the road is different from speeding while following another vehicle. Alternatively, speeding when there is a cluster of vehicles around you is even more problematic,” said Kahn.

The combination of such factors brings in differing risks associated with speeding. Apart from this, there’s also a risk associated with reaction time based on the number of vehicles surrounding the speeding vehicle. Reaction time is the time taken by a driver between when he observes an incident on the road to when he takes action – either by applying the brakes or swerving. For an average person, this takes about two seconds, and can take a lot longer for people who are impaired in any way or distracted while they are behind the wheel.  

“If you have a vehicle that is traveling 20 miles per hour over the speed limit and you couple that with a reduced tailing distance behind another vehicle, you’ve really shrunk the person's ability to react in the event of a road incident,” said Kahn. “A more dangerous scenario would be when there’s a cluster of vehicles around the speeding vehicle. In here, the driver’s ability to react or proactively maneuver the vehicle is further reduced, making it a very risky situation.”

Though speeding does not always result in road accidents, prolonged speeding does take a toll on the vehicle’s operational efficiency. For instance, a Class A tractor with a 500-horsepower engine consumes much more fuel traveling at 80 mph than at 60 mph.

Say the average lifetime speed of the tractor is 70 mph and that it averages five miles per gallon. If the tractor manages to move at its optimal speed, the mileage could be improved to 5.1 miles per gallon. If this improved mileage is compounded over a fleet of 500 such tractors, the resulting operational savings would be a significant addition to the bottom line.