Tfl Santandar Bike Analysis Calendar Heat Plot


Analysis as part of my MSc thesis, "Using Spatio-temporal Analysis and Machine Learning to predict Transport for London bike sharing habits in the post COVID-19 era".

Bike-sharing systems offer a convenient option to efficient and sustainable urban transportation, alleviating congestion. The growing demand necessitates continued understanding of how the systems are used. This study examines the shifts in Transport for London bike-sharing habits during the post COVID-19 era through exploratory and statistical analyses, focusing specifically on bike journeys undertaken in 2019 and 2022, totalling nearly 22 million rides.

The calendar heatmap reveals how a larger proportion of bike journeys occur in the summer months. Anomalies are visible in both years, but are more pronounced during 2022, such as 03/03/2022, 21/06/2022, 10/11/2022. On 10/11/2022 44,491 rides were recorded, whereas the average number of rides for all other days during November was 23,412.

September was one of two months in 2022 with fewer journeys than 2019. Noteworthy events, such as changes in the Santander Cycles fare tariff and the death of Queen Elizabeth II, likely contributed to altered transport habits and reduced bike usage during this month (TfL, 2022a; TfL, 2022b). These atypical events are difficult to incorporate into predictive models. The results indicate days with the highest bike usage are heavily influenced by TfL strikes across other forms of public transport. Notably, the top four days of bike usage in 2022 coincided with strikes on the TfL Underground system. This highlights how disruptions in other modes of transportation can impact the utilisation of BSSs. To strengthen these findings, future studies would benefit from analysing data from additional years, with a particular focus on 2023 once the data is made available. This approach would enable more robust and well-informed interpretations.

alternatetext Figure: Calendar heatmap of daily departures displayed as a percentage of the yearly total. Prepared by the author, using PostgreSQL for data storage and Python for plot visualisation.




References:

TfL, 2023a. Live Cycle Hire Updates. [Online] Available at: https://tfl.gov.uk/tfl/syndication/feeds/cycle-hire/livecyclehireupdates.xml [Accessed 1 April 2023].

TfL, 2023b. Cycle Hire Data - data format change & new data. [Online] Available at: https://techforum.tfl.gov.uk/t/cycle-hire-data-data-format-change-new-data/2520 [Accessed 1 April 2023].