Page 20 - A U4SSC deliverable - Accelerating city transformation using frontier technologies

In focus – using AI to create cycling routes, London

The City of London, United Kingdom could be considered one of the most congested functional urban areas
in Europe. Studies indicate the average time lost in traffic per year within London was 101 hours in 2015,
compared with 73 hours in Stuttgart, Germany. In order to derive a better understanding of how the roads
39
of London are being used and to create new cycling routes for the citizens, Transport for London (TfL) has
been trialling intelligent sensors that use AI to detect road users and recognise the mode of transport they
are using.
40
These sensors were deployed on a busy intersection located along Millbank and they continually gather
data on road conditions. Until the deployment of these sensors, TfL had been relying on a manual traffic
count to identify the number of cyclists on the road. The sensors have been able to monitor road condition,
identify travellers’ mode of transportation using AI 24/7 and collect valuable insights that will support TfL
in building new cycling infrastructure and reducing traffic congestion. In addition, all captured videos are
processed and discarded within seconds, meaning that no personal data are ever retained. TfL has reported
that the sensors are up to 98 per cent accurate and the city is looking to install 43 more similar sensors at
20 different locations. 41

Figure 2: Cyclists in London (Source)

The sensors are part of a wider effort by the city to modernise its road network system. It is noted that the
sensors could potentially link up with traffic signals and other control systems to provide real-time data
that will solve the congestion issue in London, which will also help reach the Mayor’s target of having 80
per cent of the journeys in London made by walking, cycling or public transport.
42

10 Accelerating city transformation using frontier technologies | A U4SSC deliverable