Challenge / Goal
The electric bike rental service in Kuopio is very popular and successful. However, the load is uneven at some stations, with certain patterns emerging that require operators to move bikes back to high-demand locations. If bikes are unavailable for rental, users may be discouraged from using this environmentally friendly option. The goal is to predict bike movements and demand, guiding the logistics operator to redistribute bikes accordingly.
Solution
To address the challenge, it was developed an AI model based on historical data to predict bike movement and demand. This model sends instructions via SMS or other communication methods to logistics operators, guiding them on where to move bikes to meet predicted demand. Additionally, a feedback loop is incorporated, where operators receive SMS queries like “Did it go well?” to provide real-time input, allowing for continuous fine-tuning and improvement of the model.
Citizen participation
Citizen participation details are yet to be finalised as the project implementation progresses.
Want to learn more about the lessons learned, financial details and results?
Log in