AI Dashcam stimulates drivers safety

Driving trucks are one of the dorsal spine not appreciated by modern civilization. It is also a difficult and sometimes dangerous job. But technology is turned towards making work safer and easier.

A new class of devices is targeted on the fleets that help drivers to escape accidents by signaling risky situations. The new systems use convolutional neural networks operating in the vehicle (“edge” AI) and in the Cloud to merge data from on -board vehicle diagnostics, as well as the data from the driver and the road. The result is systems that can assess, in real time, the risk of collision and warn motors in time to avoid most of them.

One of the most advanced new systems comes from a company called Nauto. Earlier this year, Virginia Tech Transportation Institute (VTTI) has placed the AI ​​security system compatible since the startup based in Palo Alto, California, test on the same testing roads controlled by Virginia Roads where it conducted a 2023 reference study evaluating three similar products. VTTI claims that this year’s tests were carried out in the same distracted driving scenarios, rolling stops, tail and night driving.

According to researchers from Virginia Tech, Nauto’s Dashcam has made or surpassed the gadgets previously compared in the precision of the detection – and provided comments which translate more directly into information supervisors could use to approach and correct the behavior of the driver at risk. “This study has enabled us to assess drivers surveillance technologies in a controlled and reproducible manner, so that we can clearly measure how the [Nauto] The system responded to risky behavior, ”explains Susan Soccolich, main research partner at VTTI.

The MIT driver’s attention researcher, Bryan Reimer, who has not been involved in the study, says that the real value of systems like Nauto is beyond surveillance. “Many companies focus only on surveillance, but surveillance alone is only a catalyst – the sensor, such as radar in adaptive cruise control or advanced collision warning. True art lies in support systems that shape the driver’s behavior. This is what makes nauto unique.”

Reduce fatigue alert in trucking safety

“One of our main objectives is to issue alerts only when corrective measures are always possible,” says Stefan Heck, CEO of Nauto. Just as important, he adds, is a design intended to avoid “alert fatigue”, a well-known phenomenon where alerts have triggered when situations do not really call it, potential stakeholders are less able to take into account. The false alerts have long tormented driver assistance systems, which means that drivers possibly ignore the most serious warnings.

Nauto claims that its alerts are precise more than 90% of the time, because it combines more than ten distraction and drowsiness indicators. Among the inattention indicators, the tracks of the system are the head of the head or the inclination, the yawn, the change in the flashing rate of the eyes, the long -eyed closings (indicating something called microsleeps) and the look deriving of the road for long periods (what happens when people send a text and lead). If a pedestrian enters the pedestrian crossing and the driver is awake, alert and not driving too quickly, the system will remain silent by assuming that the driver will slow down or stop so that the person on foot can cross the street without incident. But if he notices that the driver scroll through his phone, it will sound an alarm – and may also trigger a visual warning – when avoiding injuries.

Although VTTI did not specifically test the rates of false positives, it has measured the precision of detection in several scenarios. Soccolich reports that in class 8 tractor tests, the system has issued audible alerts in the cabin for 100% of portable calls, outgoing texts, discreet use on a smartphone and belt violations, as well as 95% of working stops. For the tailoring of a lead vehicle, it was alerted in 50% of the trials initially, but after adjustment, delivered alerts in 100% of the cases.

Nauto alarms can be triggered not only in the driver’s cabin, but also in the supervisors of the fleet of the trucking company that uses the system. But Nauto structures its alerts to prioritize the driver: warnings – for all situations except the most high risk – go to the truck cabin, allowing self -correction correction, while supervisors are only warned when the system detects carelessness or a lower risk behavior model which requires corrective action.

“Many companies focus only on surveillance, but surveillance alone is only a catalyst – the sensor, such as radar in adaptive cruise control or advanced collision warning. True art lies in support systems that shape the driver’s behavior. This is what makes nauto unique.” –Bryan Reimer, Mit

The company packs its vehicle equipment in a edge picking mounted on the windshield which connects to the port of diagnostic on board a truck. With the front oriented cameras and the driver and direct access to the vehicle data flows, the device continuously recalculates the risks. A delivery driver takes a look at a phone while deriving from their track, for example, triggers an immediate warning and an opinion to supervisors that the driver’s behavior justifies the carpet for their recklessness.

On the other hand, a rural stop sign in Dawn could trigger anything more than a gay reminder to stop complete next time. There are more complex cases, such as when a driver follows another vehicle too closely. On a sunny day, in light traffic, the system could let it go, retaining from the emission of a warning on the tailge. But if it starts to rain, the system recognizes the change in safe stop distance and updates its calculation of risks. The driver is invited to go back, so there is enough space to stop the truck on time on the rain road if the lead car suddenly slams on its brakes.

Nauto aims to give drivers from three to four seconds to avoid, slowly brake or refocus. “The best answer is not always slamming on the brakes,” says Heck. “Sometimes the armor is safer, and no automated braking system today will do it.”

AI Dashcams Collision Prices for lower trucking

According to a 2017 report Institute for Highway Safety (IIHS), if all vehicles in the United States were both a front collision warning with automatic emergency braking in 2014, “nearly a million rear accidents declared by the police and more than 400,000 injuries in such accidents could have been prevented”. A distinct IIHS study concluded that the implementation of the two technologies on a vehicle was sufficient to prevent half of all these collisions. Heck, pointing to these figures as well as the capacity of the Nauto system to detect the original danger outside and inside a truck, says that the compatible Dash camera of your business can help reduce the incidence of collisions even further than integrated advanced advanced assistance systems.

Damage caused by the vehicle obviously cost a lot of money and time to be repaired. The fleets also pay follow -up fees such as those associated with the drivers’s rolling, a persistent problem in trucking. Reducing crash levels, conversely, reducing recruitment and training costs and reducing insurance premiums, which gives fleet managers a strong incentive to implement technologies like this new class of AI Dashcams.

Today, the Nauto Dashcam is a supplement of spare on the size of a smartphone, but the company is considering future vehicles with integrated technology as a software functionality. Insurers increasingly fixing their prices according to the telematics of the fleets, the ability to combine video evidence, vehicle data and drivers surveillance could reshape the way the risk is calculated and the rates are fixed.

In the end, the effectiveness of these risk assessment and alert devices is the confidence of drivers. If the driver believes that the system is designed to make it a better and safer motorist rather than to serve as a monitoring tool so that the company can look over its shoulder, it will be more likely to accept comments from its electronic co -pilot – and less likely to crash.