Eyesight & Ariel U Research
Eyesight Technologies, an AI computer vision solutions leader, and Ariel University, one of Israel’s leading national research universities, today announced an ongoing collaboration on the university’s “Mobile Laboratory” (M-Lab). The collaboration is part of Ariel University’s wider study, and will provide Eyesight Technologies with unique datasets to study the state of drivers behind the wheel.
The M-Lab is an interdisciplinary research collaboration project run by experts at Ariel University in the fields of Computer Science, Robotics, and Human Experience with a focus on designing user interfaces, feedback systems, and driver-intervention schemas. The M-Lab was designed to test real-world driving scenarios by utilizing the array of sensors integrated into a test vehicle. The M-Lab researchers utilize sensors for collecting data on the road and external driving environment (e.g., Lidar and external cameras), as well as sensors for monitoring the driver’s state, such as: mental workload, detection of stress (skin conductance), vigilance (grip force on the steering wheel), fatigue (heart rate, respiration, face and eye-tracking) as well as attention on road (face and eye-tracking).
The M-Lab test vehicle is equipped with Eyesight Technologies Driver Sense, a driver monitoring system which provides the lab with accurate monitoring data related to the driver’s face and eyes. The vision layer of the company’s technology monitors eye openness, blink rate, the direction of the driver’s gaze, and head pose, among other factors to determine the state of the driver behind the wheel.
Eyesight Technologies is utilizing M-Lab’s field tests and the data from the various sensors to extract valuable insight about the state of the driver, allowing the company to further evaluate, monitor and improve its driver monitoring solution. The sensor fusion and data collaboration between all available systems provide Eyesight Technologies with crucial information beyond what is available using the driver Monitoring sensor data alone.
The M-Lab project utilizes Eyesight Technologies’ accurate vision layer, detecting and tracking head pose and a variety of eye related data, as part of the research tools used in various ongoing projects such as examining a driver’s ability to regain control of a vehicle utilizing a semi-autonomous cruise control under differing mental workloads, as well as helping to substantiate the concept of “Trusted Autonomy.”
Hyundai Mobis Detects Rear Seat Passengers
Hyundai Mobis (KRX:012330) said on March 22 that it successfully developed the system for detecting rear-seat passengers with a ‘radar’ (ROA, Rear Occupant Alert), and it is planning to propose the system to global automakers.
The ROA is a device for preventing passengers from being left unattended in rear seats. In the past, the weight sensor of children’s car seats or the ultrasonic sensor were used in general. Hyundai Mobis vastly improved detection accuracy by replacing them with the radar sensor. It is expected to greatly help prevent heatstroke accidents caused every summer by children being left unattended in vehicles and other safety accidents.
“The core of the passenger detection system, developed based on the radar which has been mostly acted as a sensor for autonomous driving, is the design of the software algorithm that discerns the micromovements of passengers,” said Mr. Jang Jae-ho, Director of the EE Research Center, Hyundai Mobis. “As the radar can penetrate clothes and measure various biosignals, e.g. micromovements of passengers’ chests and blood flow, it can detect the presence of passengers in rear seats more accurately. It overcame the weakness of the camera sensor, which cannot recognize babies covered in blankets. So far, there is no known case of application to global automakers.”
If a passenger is left in the rear seat, the passenger detection system alerts the driver with the sound made when the door is closed, or through the instrument panel or smartphone. The system, developed by Hyundai Mobis, is known to have secured electromagnetic reliability so that it works normally near high-voltage lines and railroad tracks, and be precise enough to distinguish adults, infants and pets. Hyundai Mobis is planning to develop a radar capable of measuring the heartbeats of passengers and expand the biometric function this year.