Scientists in Japan have developed a smart bicycle that can tell the difference between a rider making a normal turn and a rider losing control.
The new technology could help prevent falls by providing support only when it is truly needed, while allowing riders to stay fully in control during normal cycling.
The research was carried out by a team at Shibaura Institute of Technology and published in the journal IEEE/ASME Transactions on Mechatronics.
One of the biggest challenges for bicycle safety systems is that bicycles naturally lean when turning.
This leaning motion looks very similar to the movement that happens when a rider begins to lose balance.
Because of this, many existing stability-control systems cannot easily tell whether a rider is intentionally turning or accidentally falling. As a result, they may activate at the wrong time and interfere with normal riding.
The researchers wanted to solve this problem by teaching a bicycle to understand what the rider actually intends to do.
To achieve this, they built a special “steer-by-wire” bicycle. Unlike a traditional bicycle, where the handlebars are directly connected to the front wheel, this system uses electronic controls instead of a mechanical connection. This allows the bike to measure steering movements very accurately while still giving riders a natural steering feel through force feedback, also known as haptic feedback.
The bicycle is also equipped with artificial intelligence that continuously analyzes how the rider interacts with the bike. The AI studies several pieces of information at the same time, including the steering angle, the bicycle’s speed, how much the bike is leaning, sideways movement, and the force the rider applies to the handlebars.
By combining all of this information, the system learns to recognize different riding situations in real time.
During testing, the researchers trained the AI to identify three main situations: riding in a straight line, making a normal turn, and becoming unstable. The system was able to accurately tell the difference between intentional cornering and dangerous instability, even though both involve similar leaning movements.
This is important because the bicycle should not interfere when someone is simply turning a corner. However, if the rider starts losing balance, the system can immediately step in to help restore stability before a fall happens.
When the AI recognizes that a rider is making a normal turn, it stays in the background and allows the rider to control the bicycle naturally. If it detects signs that the bicycle is becoming unstable, it automatically activates a stabilization system to help the rider regain balance.
The researchers believe this technology could eventually be used in electric bicycles, electric motorcycles, shared bicycles, and delivery bikes. It could also be especially helpful for older adults or beginner riders, who may benefit from extra stability without losing the feeling of riding naturally.
Rather than replacing human control, the researchers say their goal is to create bicycles that work together with riders. The system acts like a quiet assistant that understands the rider’s intentions and only steps in when there is a real risk of falling.
The team now plans to improve the technology so it can recognize even more riding situations, including different road surfaces and weather conditions. In the future, smart bicycles like this could make cycling safer while preserving the freedom and enjoyment that riders value most.
