Nauto has developed a hardware-software solution capable of raising the safety level of traditional cars.
The cars of the future will be smarter and more innovative, but our daily life is still filled with traditional cars, and everything is entrusted to the driver’s attention because proactive safety features aren’t generally a reality, especially on mid-high-range cars. This is why Nauto has developed a hardware-software solution capable of raising the safety level of traditional cars.
The startup has developed a multi-sensor device mounted on the windshield and combined with two-way cameras. More importantly, the AI software running within the device is able to assess the driving environment in real-time, including the conditions of the driver, the vehicle, and the road. The platform identifies any elements of risk with real-time data, especially those that can be extremely hazardous due to the driver’s distraction.
Stefan Heck, CEO of Nauto, said Nauto leverages the same kind of deep learning-based computer vision capabilities that people are using to develop autonomous vehicles, but the company has applied them to help assist, augment, warn and improve human drivers rather than to replace them. “We basically learn from good drivers what to do, and we learn from bad drivers what to avoid; we put it into a real-time neural network that runs in the vehicle on edge. And 12 times a second it assesses the situation. If it’s an imminent collision situation, we sound an alarm and get the driver to brake or swerve. Unlike an AV or unlike AEB (automatic emergency braking), we don’t take control of the vehicle we operate through the human driver,” said Heck.
Nauto’s system provides voice guidance in the event of imminent danger with sounds that suggest the right action. So, if you need to slow down, the sound will suggest decelerating. And through this, Nauto states it can reduce risk up to 80%.
In fact, Heck reported that more than 90% of crashes involve human error and 71% of collisions are caused by distracted driving. A situation that, since 2014, has led to a 14% increase in fatal accidents and an average annual cost of between $3 and $25K per year in average cost per vehicle in a fleet.
Nauto said it is also able to warn individuals or entire fleets of cars about the safety level of both driving and roads thanks to artificial intelligence, machine learning algorithms and its cloud-based software.
The latest driver and fleet safety data show that more than 6,700 pedestrians were killed on U.S. roads in 2020, a 4.8% increase over 2019. The National Safety Council found that motor vehicle fatalities in 2020 increased by 24%, and this is the highest increase in 96 years.
Insurance costs in 2020 also increased between 10% and 15% for the third consecutive year for fleets, and driver safety is becoming an ethical and economic issue.
The technology used by Nauto provides both an inside and outside scan, thus measuring risks and identifying the relative danger as well as the proper alert to highlight. The system also analyzes drowsiness, cell phone use, and failure to wear a seatbelt. “About 50% of the risks we cover through real-time algorithms on board, another 40% of the risk we cover in a post-accident analysis in the cloud. So, today, we cover about 90% of the risk. But we want to move that 40% to the device and be able to prevent rather than just notify. We’re also looking at stop signs and red lights and visibility in reverse,” Heck said.
The company’s new advanced AI technology can track and analyze in real-time, detecting covered risks and providing preventative alerts. This is performed through computer vision telematics elements and an inertial sensor and GPS. The system is run by Qualcomm Snapdragon 845 with ARM cores.
This allows drivers to respond with sufficient time. The warning system uses more than 1.2 billion miles of driving processed by artificial intelligence to achieve high accuracy and mitigate issues related to alert fatigue from false alerts.
Heck pointed out that Nauto’s AI onboard the vehicle measures and detects the critical risks that cause collisions and helps drivers avoid them before they turn into collisions. In contrast to traditional video telematics systems that only inform managers after the fact an accident has occurred and doesn’t prevent them.
Many fleets have an anti-smoking policy, especially those that carry passengers. And we see that smoking represents a 50% in the accident category due to ash falling on the driver and the related panic caused by the situation.
Nauto recently announced an enhanced VERA Score 3.0 driver rating system based on 25 risk characteristics and is an upgraded version of its risk assessment score as part of the fleet app. Smart cameras and sensors are used to analyze these risk indicators, and VERA Score 3.0 combines all driving events and behaviors into a single score.
Each driver has a unique score based on individual journeys, ranging from 1 to 100, with 100 being the highest and safest score. The VERA scores safety across key factors and deducts points for unsafe driving. This also allows fleet managers to focus their training efforts on drivers who need it most, which is important in the resource-constrained work environments in which almost all of them operate. These insights are then provided as a dashboard and report in Nauto’s cloud application, assisting fleet management in the identification of top performers and hazardous drivers simultaneously.
The fleet app also assists with the reporting of accidents when they occur, expediting processes and clarifying fault or no-fault situations with data and video recording when the event occurs.
Nauto also offers ready-to-use reports for key performance metrics, including coaching and real-time alert efficacy, policy breaches, accidents, as well as high-performing and hazardous drivers.
This article was originally published on EE Times Europe.
Maurizio Di Paolo Emilio holds a Ph.D. in Physics and is a telecommunication engineer and journalist. He has worked on various international projects in the field of gravitational wave research. He collaborates with research institutions to design data acquisition and control systems for space applications. He is the author of several books published by Springer, as well as numerous scientific and technical publications on electronics design.