How are the bike sharing docking locations picked?
Many bike sharing systems around the world consist of two main components: bikes and the stations to dock the bikes. A bike sharing system will only be useful if the docking stations are located where people want to collect bikes and where they want to drop them off. An interesting, and very commonly used optimisation problem, is the problem of determining the best locations for the docking stations.
The problem of identifying docking station locations requires the combination of mathematics and statistics. The statistics are required to forecast the demand for bikes and identify where people will ride the bikes to and re-dock them. With the use of statistics, the expected demand for bikes can be found for all parts of the city. The optimisation problem is then formed with the constraints the ensure a minimum level of coverage of all demand is met.
The problem that is solved to identify the docking stations is the called the “Facility location problem”. This problem finds the best docking station locations that minimises the investment and achieves a minimum level of demand. In the classical problem, the docking stations could be located so that all potential riders can access them. However, this may be too expensive for the bike sharing system, since it could result in many underutilised docking stations. So this problem is typically solved to satisfy a minimum level of demand.