Optimization and Machine Learning for London Cycle Hire – Serco

Optimizing Operations for one of the largest Bike Sharing programs in the World.

Serco and DecisionBrain worked together to optimize the London Cycle Hire Scheme. With over 12,000 bikes and 800 stations, it is one of the largest city bike hire schemes in the world.

DecisionBrain developed an optimization software to assist Serco in their daily task of distributing bikes across stations, ensuring that contractual KPIs were met. Specifically, the system is designed to minimize the risk of any station either becoming empty (no more bikes available) or full (no bike drop-off allowed).

The design of the software for Serco had to be tailored to the needs of planners. Having a simple and intuitive solution was key to ensure the Serco team could quickly get up to speed and train new planners. The interface was designed to allow planners to focus on the most pressing decisions, leveraging features such as color coding, real-time status docking, status/trend visualization, 24-hour forecasting charts and more.

Record monthly hires, improved inventory, and simplified planners collaboration.

Tailoring the solution to the specific needs of the end-users and keeping complexity transparent to the planners by offering a simple and crisp UI, were critical project success factors. From a technical standpoint, combining sophisticated optimization techniques with Machine Learning Forecasting was key to achieve results.

The project was implemented in August 2017 and had measurable results by October 2017 with a record hire month, followed by a second record hire month in November. According to Serco, DecisionBrain’s solution led to very positive and quantifiable results.

• 15% reduction in parts spend
• 15% increase in driver productivity

At present, the London Cycle Hire is considered a global reference in terms of users’ experience and satisfaction. Record daily hires of 51,938 on June 24th 2020.