By Sara Miller, NoCamels -
Imagine if your car could read your mind – if you were tired, distracted, or even a little inebriated – and reacted accordingly to keep you safe.
Israeli startup CorrActions has developed software that integrates into a car’s existing system and uses hardware already in the vehicle to determine the driver’s cognitive state based on micro actions that reveal brain functioning.
Humans have a motion control center above the center of the brain called the motor cortex, which controls the body’s muscles through electric signals, explains CorrActions CEO Ilan Reingold.
As a result, he says, most of the brain’s activity is reflected in our muscles as micro motions that are unseen, unfelt and unconscious.
CorrActions takes data from motion sensors placed throughout the car that record these micro movements. The motion sensors are located in the steering wheel, where Reingold says they are very sensitive, as well as in the seats – where they were designed to register whether a child was sitting there.
“We’re taking data from these sensors and we’re analyzing it through our AI and machine learning platform,” he says.
“Through that we estimate very accurately things like the level of alcohol in the blood, fatigue and a full roadmap of additional symptoms such as stress, motion sickness and distraction.”
Once one or more of these micro movements has been analyzed, Reingold says, the machine learning algorithm in the car can react accordingly.
“According to the patterns of these micro motions, and their behavior over time, we can determine the root cause of the cognitive incapability – whether that’s alcohol or fatigue or something else – and we can determine the intensity and give it a score,” he says.
Based on the root cause and the score, the software recommends to the vehicle what kind of action to take in real time in order to compensate. These actions, such as increasing the braking distance of the car or reducing its speed, are designed to to maintain the safety envelope of the driver, regardless of their impairments.
Reingold also says that the platform can even identify the “mood” of the car’s occupants and adapt the music being played inside the vehicle to provide “a more optimized experience.”
The Tel Aviv-based startup, which was founded in 2019, has raised more than $7 million in a Series A funding round led by Volvo Cars Tech Fund and BlackBerry, which today produces advanced software for vehicles.
The objective of the founders, Reingold says, was to find a market where their technology to measure micro movements could be used to save lives. Therefore, they settled on the automotive industry, where they could detect the cognitive state of drivers in order to protect them.
The company is already working with several major car makers – although Reingold cannot say which – to integrate the software into their vehicles by 2025, for both new cars and older models that can be upgraded remotely.
“It’s a software only product – we don’t need to change any hardware in the car,” he says. “In fact, we’re in a process with car manufacturers to integrate our software in their cars through an over-the-air software update, without the need to change any hardware.”
Once the software has been installed, Reingold says, it runs without any further modifications being required.
Sign up for our free weekly newsletterSUBSCRIBE
The company is also in talks with fleet owners, both of cars and other electronic means of transport such as e-bikes and scooters.
For the latter, he says the software would be integrated into the app used to order a bike or scooter. The software would be able to determine intoxication levels based on the way in which the user navigates through the app.
In this instance, he compares the software to that of a smartphone, which has five axes for tracking the user’s finger movements and pressure.
“If their app is uploading the motion data to their cloud, and we’re running in their cloud, we can tell the mobility operator in real time if the person who’s trying to rent a scooter is drunk or not,” he explains.
Reingold says that there are other companies producing technology to register a driver’s mental state, but these technologies are “a late detection” and reactive rather than preventative.
“Some cars today use cameras that face the driver to detect if your head is nodding or your eyes are closing, in order to determine fatigue,” he says as an example. “When this happens, you’re actually sleeping, so it could be too late.”
Furthermore, the CorrActions platform is unique in that it is software based and does not require the installation of sensors that are not already present in the vehicle.
“There are other technologies, for example ignition buttons that can detect alcohol through the skin,” he says. “But again, these are hardware changes to cars, so they’re limited in what they can do in their single function solution [and] you need many of them in cars.”
Introducing new hardware to a vehicle can also take as much as five or six years to complete, while the adoption cycle for new software in cars is just one to two years.
Nor would potentially widespread introduction of driverless cars render the software obsolete. Reingold says that the automotive industry is aware that passengers in such vehicles react differently to passengers in cars with a human driver.
“One of the bad phenomena that passengers complain about [in driverless cars] is a strong feeling of motion sickness; the acceleration is not natural,” he explains.
“The turn taking and path planning is not the same as if a human driver were doing it. So the car wants to know about how passengers are affected by the driving of a robot, in order to feed this back into the automatic driver algorithms to manage the drive in a way that’s friendlier to the passengers.”
The potential for other uses of the software – such as healthcare – are not lost on the company, Reingold says, but for now the vehicle industry is the linchpin of the company.
“We’re still a small startup and we’re very much focused on the automotive and mobility spaces,” he says. “But there are many additional markets and use cases where this technology can be used and present value.”
