Illustration of the driver training assistance system (DTAS) based on triboelectric sensors (TENGs). (a) Schematic of the driver training assistance system and its components. (b) Structure scheme for the gear shift sensor. (c) Structure scheme for the steering angle sensor. (d) Structure scheme for the pedal sensor.Safe driving habits are instilled during training. Now researchers have devised a driver training assistance system (DTAS) that employs three triboelectric sensors on the gear shift, steering wheel, and pedals to monitor driving behavior in real time [Zhang et al., Materials Today (2023), http://doi.org/10.1016/j.mattod.2023.11.007].
“Improving the driving skills of drivers, particularly during the training stage, is crucial in reducing the likelihood of road traffic accidents,” point out lead researchers on the study, Xiaohui Lu, Bangcheng Zhang, Zhong Lin Wang, and Tinghai Cheng. “The DTAS proposed in this work aims to solve the problems of time-consuming and laborious traditional training methods and low training efficiency.”
The team from Beijing Institute of Nanoenergy and Nanosystems, Changchun University of Technology, Jilin University, and Georgia Institute of Technology harnessed triboelectric nanogenerator (TENG) sensors developed by Zhong Lin Wang’s group as the basis of the system. These self-powered copper electrode-PTFE film devices, which harvest energy from the environment, were placed in different areas of a driving simulator to monitor all aspects of a driver’s behavior. The TENG sensors provide data on the selected gear, the direction and turning angle of the steering wheel, and whether the pedals are being pressed or released, which is collected by an electronic reading device.
However, since the data from TENGs is affected by many external factors, the information is fed into a deep learning convolutional neural network technology, which can process and evaluate the data in specific scenarios. One-dimensional convolutional neural network technology is particularly effective in extracting subtle features from time series data, point out the researchers. In combination with deep learning, the DTAS can analyze and indicate whether driving is meeting the required standard for a given scenario, such as reverse parking, with 97.5% accuracy. Information from the DTAS, moreover, can be relayed to the driver in real time through a display interface on the dashboard.
Compared with traditional video monitoring of driver behavior, the TENG-based DTAS has lower equipment costs, is easy to install, provides immediate feedback, and is less intrusive for the driver. The researchers believe that their proposed DTAS could assist in the training of novice drivers and has the potential for new models of driver training without instructors.
“The novelty of this work is that the triboelectric sensor… is not only simple in structure and easy to install, but also avoids the problems of power supply and complex wiring in the limited space of the vehicle,” say Lu, Zhang, Wang, and Cheng.