# Can Steering Wheel Detect Your Driving Fatigue?

**Authors:** Jianchao Lu, Xi Zheng, Tianyi Zhang, Michael Sheng, Chen Wang, Jiong Jin, Shui Yu, Wanlei Zhou  
**arXiv:** [2010.10327](https://arxiv.org/abs/2010.10327)  
**Date:** 2020-10-18  
**Summary page:** https://www.summarizepaper.com/en/arxiv-id/2010.10327v1/

## Key points

- Automated Driving Systems (ADS) have potential to enhance safety, mobility, and efficiency of transportation systems
- Current ADS design employs human-in-the-loop (HITL) to monitor driver behavior
- Existing methods for detecting driver fatigue are not always reliable
- Authors propose a novel method using surface electromyography (sEMG) sensors on a steering wheel to detect driver fatigue
- Advantages of this approach include non-intrusive collection of bio-signals and earlier detection of fatigue
- Experiments show the proposed method outperforms other approaches with weighted average F1 scores of approximately 90%
- Future directions involve incorporating supplementary sensors for multimodal learning to enhance accuracy and robustness
- Proposed method addresses limitations of existing approaches and offers non-intrusive means to collect bio-signals and detect fatigue at an earlier stage

## Summary

Automated Driving Systems (ADS) have gained significant attention in both industrial and academic communities due to their potential to enhance the safety, mobility, and efficiency of transportation systems. The current state-of-the-art ADS employs a human-in-the-loop (HITL) design, which closely monitors the driver's behavior for any anomalies. However, existing methods for detecting driver fatigue primarily rely on vehicle driving parameters and facial features, which are not always reliable. In this paper titled "Can Steering Wheel Detect Your Driving Fatigue? ", authors Jianchao Lu, Xi Zheng, Tianyi Zhang, Michael Sheng, Chen Wang, Jiong Jin, Shui Yu, and Wanlei Zhou propose a novel method for detecting driver fatigue by embedding surface electromyography (sEMG) sensors on a steering wheel. This approach offers several advantages over existing methods. Firstly, it allows for non-intrusive collection of bio-signals from the driver. Secondly, it enables the detection of driver fatigue at an earlier stage. The authors conducted experiments to evaluate the performance of their proposed method and compared it with existing approaches. The results demonstrate that their approach outperforms other methods with weighted average F1 scores of approximately 90%. These promising findings highlight the effectiveness and reliability of using sEMG sensors on a steering wheel for detecting driver fatigue. Additionally, the authors suggest future directions for deploying this approach in real-life settings. One such direction involves applying multimodal learning by incorporating several supplementary sensors. This integration could further enhance the accuracy and robustness of driver fatigue detection. Overall, this paper presents a significant contribution to the field of ADS by introducing a novel method that addresses the limitations of existing approaches for detecting driver fatigue. The proposed use of sEMG sensors on a steering wheel offers a non-intrusive means to collect bio-signals and detect fatigue at an earlier stage. The high-performance results obtained from experiments validate the effectiveness of this approach, and the proposed future directions provide valuable insights for further research and implementation in real-world scenarios.