With the advent of autonomous driving, it’s becoming increasingly important to understand the driving behavior of humans. LabShare’s partner, Nervtech offers all the complex services and technical solutions this task requires. But how does human behavior become measurable?
Digitalization allows us to continuously improve and optimize the technologies we use, and the automotive industry is no exception. An increasing level of automatization gives way to a safer, predictable and accident-free means of transportation.
But when it comes to road safety, information and communication technologies (ICT) can help us even further - partly by making it possible to analyze drivers’ behavior like never before.
Research focusing on drivers is gaining popularity, supported by improvements in machine-to-machine (M2M) communication, as well as the availability of wearable sensors. But how are these devices important to monitoring human behavior?
Vehicle automation is rapidly evolving, but in order for these vehicles to be introduced into practical use, we need to authentically imitate human drivers. This requires measuring their moves, actions, decisions, different behaviors and the correlation between them. Various algorithms need to be developed based on these results, for which driver profiling is required.
To this end, we first need to understand the responses of the human body to specific driver’s states and behavior. Which psychological signals are the most important and how can they be measured?
Analyzing a driver’s movement
The signals that represent drivers’ state are connected to the cardiovascular, digestive, immune, muscular, nervous and respiratory system of the body. Therefore, the key tests look at signals connected to these systems.
Electromyography (EMG) can detect the electrical activity of the muscles, particularly leg muscles that press clutch, break and throttle pedals. Therefore, EMG can be used to detect how much effort it requires from the driver to use the pedals.
Eye tracking, as the name implies, refers to tracking eye position in real time, calculating gaze points and measuring other eye parameters. In setups with controlled brightness, cognitive load can be extracted from the pupil diameter.
Accelerations are most often the consequence of body movement. Accelerometers can also be used to detect chest movement from which breathing patterns can be extracted. Spirometry primarily measures the speed and the volume of air a person inhales and exhales. Lung capacity and one’s ability to handle stress can be concluded from spirometry results.
The autonomic nervous system and brain activity
Not only visible physical signs, but also some other processes of the human body can be measured - even the way our brain functions in a given moment.
Reactions of the nervous system, such as increased or decreased activity, normal or abnormal activity are observed through electrocardiography (ECG) that measures the electrical activity of the human heart during contractions.
Cutaneous electrogastrography (EGG), on the other hand, measures of gastric slow waves and can be correlated to gastric motility. This test is performed with electrodes placed on the abdominal skin.
Finally, electroencephalography (EEG) measures signals that directly represent brain activity: from these readings, components of brain waves can be extracted. Alpha waves are expressed when persons have their eyes closed or are relaxed. With the opening of the eyes or excessive mental workload, beta waves become more noticeable. By detecting alpha waves, EEG can successfully measure a driver’s drowsiness and relaxed state, while with the presence of beta waves, a driver’s cognitive load can be measured. EEG measurement requires a special helmet with many electrodes and wires.
How to interpret these results?
Measuring different psychological signals during driving can help us determine a driver’s state and behavior. This does not only provide a clear overview of the condition a driver currently is in, it also allows us to make accurate predictions regarding future behavior. The result? Safer, more efficient and accident-free transportation that ever before.
From the signal captured with digital devices, we first have to determine what kind of behavior is typical for individual driver profile, and then try to profile drivers based on observations of their behavior. Insurance companies already use driver profiling, mostly to detect risky drivers, but the results can also be used to provide drivers their driving scores for self-evaluation and training.
Would you like to find out more about this exciting new field, or even organize driver profiling for your company? Have a look at Nervtech’s website or get it touch with them via LabShare’s platform!