Lynn Walford, AUTO FUTURES
Cameras watching roads and the faces of drivers are currently proving to be very helpful when it comes to safety and security. But camera video can also act as witnesses and provide data to trigger alerts that warn drivers that they are distracted or drowsy. And, in the near future, cameras can be used with facial algorithms to start the car or even pay for fuel.
Witness Protection
Video recordings from automotive cameras of the road and drivers provide protection and training opportunities for commercial businesses and drivers.
“It is important to have cameras in fleet vehicles for insurance claims and to protect liability,” says Andrea Mirchef, Account Manager at SmartWitness USA, a provider of automotive video camera systems. She says it’s important that the cameras are hardwired into the vehicle so that drivers don’t remove or disable them.
Video and data help to defend against insurance fraud and increased premiums. Even individual drivers can save on the cost of car insurance.
“After I had the SmartWitness camera system installed, my car insurance went down $10 a month,” confirmed Mirchef. She noted that SmartWitness video proved to her insurance company that she was a victim of a hit and run accident with a big rig truck.
SmartWitness also provides a software API for third-party telematics and GPS providers as well as front and driver facing camera systems. SmartWitness driver behaviour analytics and smartphone apps help to improve driving skills.
Cameras That Wake Up Distracted or Drowsy Drivers
“Driving while distracted is an epidemic problem,” said Kevin Tanaka Senior Director of Marketing for SeeingMachines, a company that provides driver monitoring system software, processors and AI for use by Tier 1 suppliers to automakers.
SeeingMachines’ FOVIO driver monitoring processor and technology are in the 2018 Cadillac CT6 for use with semi-autonomous Super Cruise. Head movement is tracked and the system alerts drivers to pay more attention or take back control.
SeeingMachines uses NIR (Near Infrared) LEDs with camera sensors and a filter that can see through most sunglasses, sun glare and poor lighting conditions. The software monitors several points on a face for eye gaze, blink rates and pupil diameter to determine drowsiness or distraction.
“Eye tracking allows us to understand the direction of the eyes as well as how long someone is looking at something. It can tell if someone is looking for something for four seconds like at a phone,” says Tanaka.
SeeingMachines started in the fleet space for commercial driving. Its Guardian systems monitor driving, stream video and connect to a control centre.
Guardian has observed 1.3 billion kilometers and video of millions of drowsy events which can show unexpected occurrences such as video of a driver who was asleep with his eyes open. Tanaka says that fleet instances of driving while distracted or drowsy have been reduced by 90% through Guardian use.
In the case of a totally non-responsive driver, the automaker can then deploy data to a call centre, emergency responder system or another safety system.
Besides Cadillac, SeeingMachines is working with two German automakers and has a signed agreement with a second US automaker.
SeeingMachines also offers a backup system to monitor human autonomous backup drivers, called Guardian Backup-driver Monitoring System (Guardian BdMS).
“It is very hard for the driver to continue to pay attention and it gets boring when the car is doing the driving,” says Tanaka.
Guardian BdMS has an audio warning, light warning and can be attached to a haptic seat unit that vibrates the seat to wake up the driver.
Many European automakers will be installing driver monitoring systems. By 2020 vehicle manufacturers wanting to qualify for a full Euro NCAP safety rating will have to have a driver monitoring system (DMS) onboard.
Employing Cameras for Facial Recognition
In the future, cameras may also be used to identify the driver and act as a security measure; a bit like FaceID is used to authenticate iPhone users
“Your key fob or key could be stolen but criminals can’t copy your face,” says Kate Migon, Head of Automotive and Mobility Services Americas at Gemalto. Gemalto provides facial recognition algorithms that currently are being used for border patrol, airports and casinos that cannot be spoofed.
“Facial recognition doesn’t require extra effort such as a PIN or another device. It’s passive. It’s not something you have to do extra,” adds Migon.
Possible use cases for facial recognition include who can start the vehicle and when. A family member such as teenager may only drive the car to school and home, not late at night. Or the car may only play family-appropriate music when a child is detected. When cars have payment systems installed in the vehicle only the face of the credit card holder will be able to purchase fuel, food or merchandise through the onboard payment system.
But, no matter what happens in automobiles in the future, drivers can’t let all the new technology distract them. Experts say the driver has to pay attention as they are ultimately responsible.