Autonomous Bicycle Brake Assistance Using Dynamic Grip Force Measurement

Bicycle accidents, particularly those involving elderly cyclists, pose significant societal and financial risks. The ineffective operation of brakes is a crucial contributing factor to these accidents. This paper proposes a novel autonomous bicycle brake assistance system to enhance braking performance for users with limited grip force or slower reaction times. Our system comprises three main components: an assist force calculation mechanism, a brake force assistance system, and a dynamic assistance adjustment mechanism. Experiments conducted with elderly participants demonstrated significant improvements in braking distances, with assisted braking closely aligning with target values. Furthermore, the system successfully detected and adapted to changes in user performance under fatigue conditions using the Local Outlier Factor algorithm. These results suggest that our proposed system can effectively enhance bicycle safety for vulnerable user groups by providing personalized, adaptive brake assistance.