Many traffic accidents are caused by drowsy driving or driver negligence.
By the time the driver becomes aware of this, he must have already lost his concentration. Is there any way to inform this in advance?
This is why Hyundai Mobis developed the world’s first brain wave-based driver monitoring system.
When are you most sleepy during the day? The answer will be different depending on each situation, but most people say it is after meals. You cannot avoid drowsiness while driving. Fatigue from long driving is also a factor that threatens safety. In this case, the average driver would open the window to refresh the cabin, then stop in a safe place to overcome drowsiness. However, the situation is not easy for bus drivers who have to go around a set route within a certain amount of time.
As a result of a study conducted by the Gyeonggi Research Institute on 140 public bus drivers, not only individual conditions such as sleep time but also working hours had an effect on drowsy driving. In particular, they confirmed that not having enough rest after driving or driving after eating had a significant impact on drowsy driving.
Drowsy driving or negligence is one of the major causes of major traffic accidents. Automakers are applying driver negligence monitoring technologies to prevent this type of traffic accident. The driver’s attention is assessed by combining the driving data such as lane departure, steering wheel control, acceleration and deceleration, and the other data of driver status such as eye blink detection. If the system detects a problem, it warns the driver with an audiovisual message.
However, the technology has fundamental limitations. It only detects problems after the driver’s attention has been reduced. Reducing the chance of an accident requires a new level of safety technology – warning car anomalies before they happen. This is why Hyundai Mobis developed M.Brain.
M.Brain: The world’s first brainwave-based driver monitoring system
Hyundai Mobis’ M.Brain is the world’s first brainwave-based driver monitoring system. It analyzes the driver’s brain waves to check health conditions such as fatigue, drowsiness, and distraction to warn him. It can also be linked with a smartphone app.
For reference, EEG is considered one of the most challenging biosignals to analyze. However, the size of the M.Brain is similar to that of a small earphone. Miniaturization minimizes the inconvenience of drivers who have to wear them all day long. Compared to the current brain wave sensor, it is small enough to be called an innovation.
Reduced size, higher signal quality
The most difficult part of developing M.Brain for Hyundai Mobis was reducing the size and improving the signal quality at the same time. In general, small earphones are comfortable to wear and easy to use, but the signal quality may be poor.
Hyundai Mobis focused on removing high-frequency noise and motion noise (noise caused by motion) to ensure the best signal quality. Brain waves, which are electrical signals, are so delicate that even the blink of an eye creates a motion noise. Therefore, it is essential to remove the interference caused by the motion noise and to secure a high-quality signal. Hyundai Mobis used the LMS (Least Mean Square) algorithm to find optimal coefficients to remove motion noise and verified the function using Simulink, enabling graphic-based dynamic system data analysis.
Machine Learning to Increase Accuracy
A large amount of data is required to determine whether the driver’s attention is reduced by brain waves. However, in the case of biosignals, it is complicated to build a database with only desired signals and obtain high-quality signals for learning. Therefore, many companies mainly use data augmented learning techniques to solve this problem. It is to add multiple transformations to the underlying data to create various examples and use them for machine learning.
Hyundai Mobis used a data augmentation technique using geometric transformation; For example, they created variations through techniques such as ‘jittering’ that adds irregular noise to the signal, ‘scaling’ that increases/decreases the signal, and ‘stretching’ that expands/reduces the time axis – and then used them in machine learning. So they were able to increase the accuracy of the EEG interpretation.
M.Brain: Making Buses Safer
M.Brain is currently being introduced to Gyeonggi-do public buses as a pilot operation, helping to create a safer public transportation environment. Gyeonggi Province, Gyeonggi Research Institute, and Hyundai Mobis have signed a business agreement to distribute M.Brain to public bus drivers. M.Brain plans to supply 300 units in 2022, starting with 10 units in 2021, and then gradually increase the number every year to improve technological perfection.
Despite being a pilot program, M.Brain is already showing noticeable effects. According to the 2021 data from Gyeonggi Research Institute, when driving a bus while wearing M.Brain, negligence in driving decreased by 25.3%. In particular, it showed a high reduction effect of 29.7% when driving after a meal. M.Brain also appears to impact attention span: from 6.7 seconds without alarm to 2.3 seconds with a warning on. Hyundai Mobis plans to continuously improve M.Brain from various perspectives to increase satisfaction.
A new possibility: a car that reads the mind of the driver
M.Brain pioneered a new area in the automotive healthcare market. It is a technology that uses brain waves, which is the most difficult to understand among biosignals. This is because brainwave-based technology can develop in various directions as the amount of measurable data increases. Hyundai Mobis plans to secure various databases for brain wave signal interpretation and software logic development through machine learning.
As such, M.Brain holds various possibilities. It is now used to help bus drivers to keep everyone safe. Still, in the era of autonomous driving, it can identify abnormal passenger status or develop into other safety features. It is also possible to understand the psychology of the occupant by combining other bio-signals. So it could evolve into a technology that controls cabin scents, lights, driving style, etc. – the era of ‘cars that can read your mind,’ indeed. This is why we should pay attention to the future mobility technology of Hyundai Mobis.