Static vs. Dynamic Rebalancing — Why Static Doesn’t Work
The biggest challenge in bikeshare systems is to provide the best possible service each and every day, no matter the weather, season, resource availability or ridership.
Operators that choose schedule-based distribution are often limited in their rebalancing efforts and are not efficiently serving the user demand, this impacts both ridership and user experience as a whole.
It’s up to the operators to ensure they have a simple and optimised management process that’s tailored to their cities and users. It enables them to offer better services, build stronger ridership and reduce operational costs. Yet, many operators still choose to carry out schedule-based rebalancing.
Schedule-based distribution means drivers have pre-set information or a schedule of how many bikes to pick-up and drop-off and at what stations or locations, sometimes dependent on historic traffic patterns. The schedule is created irrespective of the prevailing conditions of the city such as weather, traffic, transport strikes, available bikes or operation team’s availability. This can often mean that by the time the first job is completed the whole schedule could be wrong and the operator is playing catch-up.
Poor rebalancing and management processes can also add to the growing situation of bikes being left as a nuisance to cities and its citizens. With the very nature of bikeshare being dynamic, a static based approach is now of the past.
It is now more important than ever to have a smart management process that takes the guess work away from bikeshare rebalancing.
Here’s four reasons why schedule-based rebalancing doesn’t work for modern Bike Share Schemes:
1. Dynamic Nature of the City
Cities globally are different from one another. They all have different population, city topography, transport hubs and many other factors that make each and every city unique. A schedule that works well in one city could completely fail in another.
2. Each Day is Different
The day itself plays a huge role in bikeshare ridership. When you combine the changing weather, major city events, transport strikes and a number of other things, the demand for bikeshare systems are likely to fluctuate on a daily basis.
3. Evolving Rider Behaviour
Some riders may take a bike out every morning and every evening to commute between work and home. That makes it predictable and easy to manage. In reality, for many riders their behaviour is constantly changing. Increase in tourists or a rise in public transportation services could affect the demand for bikes in different areas.
4. Inefficient Use of Staff & Resource
What we see every day with schedule-based distribution is that drivers are going from one location to another with no real insights. They often go to a docking station that is expected to be empty and find that it is full. This is huge waste of time, money and resources.
For modern bikeshare systems, scheduled-based rebalancing doesn’t work. There are too many variables that make schedule-based rebalancing time-consuming and inefficient. Operators instead need simpler and smarter processes that can predict demand and manage resources with accuracy.
At BICO AI, we use four weeks of prior system data to track rider behaviour and manage rebalancing effectively. We predict optimal fill values up to 24 hours in advance per station whilst assessing prevailing conditions and resource availability to enable operators to make quick decision and move quickly to win and maintain new riders.
BICO AI provides an AI Operations Platform that assesses various environmental, resource, performance and usage data to optimise available resources (teams, bikes, batteries etc.) to ensure asset availability (location, usable, charged) across the entire network when and where it is required.
