Intelligent Bicycles: Using AI to Improve Cyclist Safety & Reduce Injuries

Votes: 0
Views: 305

Every year in the US, according to NHTSA, there are on average ~750 cyclist deaths and ~48000 injuries due to crashes with motor vehicles. Total economic impact is estimated to be >$10 billion annually aside from emotional and psychological suffering of affected families. Distracted/compromised motor vehicle drivers & inattentive cyclists are among the top causes of the crashes based on information from PPM (People Powered Movement).

By combining AI (artificial intelligence) with the latest sensing technology including radars and cameras, the proposed intelligent bicycle design is intended to significantly reduce cyclist deaths & injuries due to crash with motor vehicles. Specifically the sensor data collected including speed, direction, and traveling patterns of both the motor vehicle and the bicycle would be analyzed by the AI engine. A collision probability will then be calculated, and visual and audio warning signals deployed in a tiered strategy to get the attention of the driver & the cyclist. For example, if the level is determined to be low, only visual signal through slowly flashing neutral color light such as blue will be used. For high risk situations, one can utilize fast flashing red light in combination with urgent audio warnings.

One can further expand the value of this technology by developing an App to track the cyclist's riding patterns. The information can then be analyzed to provide feedback to the cyclist to improve riding experience and enhance health benefit. A community of riders using this technology can also be formed to enable information sharing.

The proposed design is feasible technically with existing sensing technology and AI capability. With ~20 million bicycles sold per year in the US with total annual value of $9.5 billion that is projected to grow to $14 billion by 2030, the market potential for such a product can be quite large in the US alone.

Voting

Voting is closed!

  • ABOUT THE ENTRANT

  • Name:
    Steven Jia
  • Type of entry:
    individual
  • Patent status:
    none