Powered by BICO: Changing Patterns

The impact of COVID-19 on bikesharing operations has initially been mixed with the longer term outlook looking relatively positive.

While the face of bikeshare systems will likely look the same, operations will undoubtably look very different as we enter the ‘new normal’.

As bikeshare operators begin to emerge from closed systems and reduced ridership, it is proving that the past is not a good indicator of the future.

Users are now taking longer journeys and are also starting and ending their rides at different locations then previously observed. These changes are only further compounded by an increased number of users, both new and returning.

So, what happens when the goalposts move, and your operational playbook becomes redundant?

Working with bikeshare operators across the industry has given us a unique insight into overcoming not only the daily challenges of bikeshare operations but also the new challenge of returning from lock-down. One of these challenges is change in usage patterns i.e. where users ‘traditionally’ start and end their bikeshare trip.

Rider usage patterns are seen as a critical data point to enable demand forecasting of system usage. By understanding where and when users will be renting bikes (starting their journey) and parking their bikes (ending their journey) operators are able to predict with a degree of accuracy of what the optimal fill level of a station should be to meet this demand.

The level of prediction can vary from what an experienced field team member assesses is the ‘normal’ pattern based upon experience to excel spreadsheets that analyse these patterns on a monthly to quarterly basis to set optimal fill levels for the rebalancing teams in the field. However, when a transport strikes or global pandemic occurs these patterns not only change significantly in the immediacy but also remain different (and at times continuously changing) after the event.

Therefore, it is critical to use a data-driven approach that accounts for subtle changes in key parameters over a short timeframe to ensure that bike and parking availability is available at the stations where it is required.

So, What Does This Mean?

As the world remobilises from lockdown caused by COVID19, it is evident that user behavioural patterns have begun to change quite dramatically.

The change is likely due to cycling being seen as the safest transportation mode during the crisis, along with various membership offers being introduced by systems such as free rides for the essential workers and reduced costs of rides.

These changes affect the system and subsequent operation by seeing distinct changes in higher usage stations and longer journeys, as observed in London. This requires operators to reassess their ‘Priority Stations’ and therefore operational focuses across the entire system.

Systems utilising our BICO system have not had to focus on these changes, as BICO automatically assesses these changes in real-time and makes autonomous decisions to best rebalance the system in a data-driven way.

By automatically harvesting various data points regarding the system usage, BICO calculates usage changes, patterns shift, bike & resource availability, KPI performance and optimal fill levels to provide optimal rebalancing strategies for the system in real-time.

For example, the largest European docked based operation outside of China who utilises the BICO system, were enabled to transition for drastic reductions in ridership, usage patterns and operational resource to a surplus of demand that broke all previous records without having to complete any analysis or change their operations strategy due to it being completed intelligently and autonomously by the BICO system.

It also allowed them to shift strategic priority focuses instantaneously by moving ‘Priority Stations’ for high usage areas such as transport hubs and tourist attractions to hospitals and other essential areas with a click of a button. Thus ensuring 99% usability not only at critical stations but across their entire bikeshare network.

Which is why it is critical, based upon the operational, financial and usage impacts caused by the pandemic that bikeshare operators need to decide:

“What does good look like?”

Many bikeshare systems have no explicitly defined purpose, with existing rebalancing strategies supporting or clashing with the purpose or suggested benefits of a bikeshare system.

Combined with ever changing parameters such as usage patterns, resource availability and ridership — this is an ever evolving problem (and headache for operators)

By understanding what good looks like and utilising a system or strategy that autonomously manages the dynamic nature of bikeshare systems, operators further improve their level of service to the user and city (and outperform their KPI’s)

Ultimately, utilising a data-driven approach to changing usage patterns gives operators a slight advantage to pro-actively manage and prepare the system(especially overnight) to meet the demands of the coming day/rush-hour. However, mis-calculations due to subtle changes in usage patterns can have a disastrous impact.

Where reaction to change is often too late, a pro-active, data driven approaches are essential to enable optimised system performance.

Read Part-1 of the Powered by BICO Series: Workforce

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.