This technology focuses on self-balancing and auto-driving of electric bicycles. The technology mainly contains four items.  The first item is the sensor fusion method that estimates the bicycle＇s posture in real time.  The second is the robust control algorithm that allows the servo motor to control the handle to achieve self-balancing and steering. The third item is positioning technology which combines GPS with motion model estimation to accurately and instantly compute the coordinates and orientation of the bicycle. The fourth item is obstacle detection and obstacle avoidance path planning technology based on deep learning. By the integration of these four items, electric bicycles can achieve autonomous driving.

The technology mainly contains four items.  The first item is the sensor fusion method that integrates gyroscopes, accelerometers and the wheel encoder to estimate the bicycle＇s posture in real time.  The second item is the robust control algorithm that allows the servo motor to control the handle to achieve self-balancing and steering. The third item is positioning technology which combines GPS with motion model estimation to accurately and instantly compute the coordinates and orientation of the bicycle. The fourth item is obstacle detection and obstacle avoidance path planning technology based on deep learning.  By the integration of these four items, electric bicycles can achieve autonomous driving.

Taiwan is an internationally well-known kingdom of bicycle manufacturing, and there are world-class manufacturers such as Giant and Merida. In recent years, due to the establishment of the U-bike system, people rely more on bicycles for commuting, sports and leisure in the city. The self-driving bicycle technology established in this study can improve Taiwan’s technological leading position and brand-name image in the international bicycle industry.  When applying it to the U-bike system, the U-bike can be like the driverless taxi that automatically drives to the rider and leaves by itself when reaching the destination.  Moreover, the auto-driving technology can be implemented as the assisted driving mode so as to increase the intelligence and safety of bicycle riding.