Optimising London’s bus services and network

As part of the RILM Project, Transport for London (TfL) wanted to optimise London’s bus network and timetables, supporting The Mayor’s Transport Strategy. This strategy outlines various goals, including having 80% of London trips made by foot, bike or public transport by 2041.

To help achieve this goal and optimise public transport services, TfL specifically, wanted to understand:

• Bus travel time variability, so where buses experienced congestion, impacting travel times.
• Traffic signal optimisation, and arrival on green, so were buses free-flowing or experiencing split failures. (A split failure occurs when a traffic signal does not provide enough green time to allow previously stopped vehicles to continue through the intersection, making them wait for longer than one cycle.)

Compass ingested, road-matched, and analysed TfL’s bus data so it could be visualised on custom dashboards, and combined with general traffic data for better intersection analysis. This allowed TfL to understand location, speed, journey times and direction of travel for specific buses and route numbers.

Compass’ Data Science team built a dashboard analysing the bus data across the bus network. The dashboard was filterable by date range, day of week, peak period, intersection number and name, intersection leg, and bus route. Within the dashboard, TfL could see how all vehicles (i.e. cars, HGVs, LGVs as well as buses) behaved on bus routes; or just isolate to buses.

Looking at a sample of data (18-19th Feb), TfL wanted to see where select bus routes (113, 13, 94, 2, 137, 134, and 188) were experiencing split failures. Results identified multiple intersections, with additional insights such as:

• Average waiting time of up to 4.4 minutes at some intersections.
• A maximum wait time of 5 minutes recorded at one intersection in the AM Peak.
• Average travel time through intersections ranging from 7.53 seconds to 4.4 minutes.
• A maximum travel time of 5 minutes recorded at one intersection in the AM peak.
• A maximum queue length up to 323.78 metres recorded at one intersection in the off peak.

TfL are using this dashboard upscale their investigation, understanding where to enhance bus service reliability and optimise signal timing, reducing average journey times across key intersections in London.

This use case is part of the RILM project supported by EIT Urban Mobility, an initiative of the European Institute of Innovation & Technology (EIT), a body of the European Union. To read more about the RILM project, visit https://www.compassiot.com.au/uk/ri-for-london-mobility.