科技科学

Self-driving cars make mistakes, and now users can see why

Nature正文已收录本站
  • NEWS AND VIEWS

An autonomous-vehicle system that directly explains its decisions in a way that can be interpreted by humans could improve driving safety.

By

  1. Hyochang Kim

    1. Hyochang Kim is at the Stanford Center at the Incheon Global Campus, Stanford University, Incheon 21985, South Korea.

  2. Hyunmin Kang

    1. Hyunmin Kang is in the Department of Psychology, Daegu University, Gyeongsan 38453, South Korea.

An autonomous vehicle brakes on a seemingly empty road. The driver cannot tell whether the car has detected a real hazard or is hallucinating and, unable to understand the system’s reasoning, they do not know whether to intervene. Writing in Nature, Kenny et al.1 report an algorithm called Concept-Wrapper Network (CW-Net) that tackles this ‘black box’ problem. CW-Net uses human-interpretable concepts to make the vehicle’s decisions transparent, and the authors show that this makes people better at predicting what the vehicle will do next — including in situations in which it has made a mistake.

References

  1. Kenny, E. M. et al. Nature 657, 114–120 (2026).

    Article  Google Scholar 

  2. Rudin, C. Nature Mach. Intell. 1, 206–215 (2019).

    Google Scholar 

  3. Koh, P. W. et al. in Proc. 37 Int. Conf. Mach. Learn. (eds Daumé, H. III & Singh, A.) 5338–5348 (PMLR, 2020).

    Google Scholar 

  4. Endsley, M. R. Hum. Factors 37, 65–84 (1995).

    Article  Google Scholar 

  5. Koo, J. et al. Int. J. Interact. Des. Manuf. 9, 269–275 (2015).

    Article  Google Scholar 

  6. Lee, O., Currano, R., Miller, D., Kim, H. & Sirkin, D. in Proc. 16 Int. Conf. Automot. User Interfaces Interact. Veh. Appl. 248–258 (ACM, 2024).

    Google Scholar 

Download references

Competing Interests

The authors declare no competing interests.