AutoBike's Self-Riding Bicycle Uses Hesai Lidar to Explore Autonomous Bike-Sharing Redistribution
Ten ETH Zurich students turned a standard bike into an autonomous prototype with a Hesai JT128 lidar, proving it can ride itself across campus and park.
- up to 30% of revenue
- 75%
- 10 students
- ~9 months
What Happened
Bike-sharing networks struggle to keep stations balanced as commuters empty some hubs and fill others, forcing operators to move bicycles manually with vans and staff at up to 30 percent of revenue. AutoBike, a focus project by ten bachelor students at ETH Zurich's Autonomous Systems Lab, explored an alternative: a bike that rides itself to where it is needed. In roughly nine months, the team converted a standard bicycle into a working autonomous prototype and completed a proof-of-concept ride across the ETH campus.
75%
Reported by the AutoBike team on the final day of testing.
The prototype keeps the look and function of a normal bicycle while hiding the systems that make autonomy possible: hidden steering, retractable support wheels, onboard actuation, a hydraulic braking system, a camera, and a Hesai JT128 lidar. The single lidar builds a 3D point cloud of the environment, giving the bike direct distance, shape, and position data to localize itself, detect obstacles, identify free space, and plan a path in real time. Because lidar measures distance directly, perception is less dependent on ambient light, supporting operation from daylight to darkness.
- Hidden steering
- Retractable support wheels
- Onboard actuation
- Hydraulic braking system
- Camera
- Hesai JT128 lidar sensor
AutoBike completed its first autonomous ride on the ETH campus, driving from the PubliBike station at ETH across the Polyterrasse and autonomously parking at the station near the university. During the ride, the Hesai JT128 lidar captured the environment as a 3D point cloud, which AutoBike used to understand its position, identify obstacles, detect open space, and calculate a safe path. The ride demonstrated the concept's technical credibility.
AutoBike is not yet ready for public bike-sharing use; the team says further development and testing are required before operation in urban environments. Next steps under consideration include potential pilot testing, further system development, and cost reduction. For Hesai, the project illustrates how one lidar sensor can support autonomy on compact platforms where precise perception is central, not an optional enhancement.
Previously from Hesai Technology
This is not Hesai's first push beyond autonomous driving. In prior coverage, Hesai argued that 3D lidar is finding new value in the final, slow-motion meters of docking operations, from aircraft gates to robot chargers. The AutoBike proof of concept applies that same lidar-driven precision to a bicycle parking itself at a docking station.
Background drawn from MotorClaw's earlier coverage of Hesai Technology's official releases.
Why this matters
Bike-sharing operators often pay up to 30 percent of revenue to manually rebalance stations with vans and staff. AutoBike's proof of concept suggests that bikes with reliable 3D perception could eventually ride themselves to underfilled stations, cutting that cost and improving bike availability for users. The project also shows lidar's value extends beyond cars to compact, dynamically balanced platforms.
Terms in This Story
- lidar
- A sensing method that uses laser light to measure distances and create a 3D map of the surroundings.
- point cloud
- A set of 3D points that represent the surfaces of objects in an environment, generated by lidar.
- proof of concept
- A small-scale demonstration that an idea or technology can work in practice.
- bike-sharing
- A service that offers bicycles for short-term public use, usually through docking stations.
Summarised from the linked release; details can be imperfect — always verify against the original source.