Aims of the project:

  • Using low power, low cost hardware and the latest machine learning algorithms and computer vision research Vivacity Labs are deploying a network of intelligent sensors, able to monitor real-time transport at city-wide scale. In the VivaMK project, Vivacity Labs are giving Milton Keynes Council the unprecedented ability to understand, model and optimise multi-modal transport networks, while also providing citizens with next-generation journey planning data, such as the quickest route to an empty car parking space.

Using low power, low cost hardware and the latest machine learning algorithms and computer vision research Vivacity Labs are deploying a network of intelligent sensors, able to monitor real-time transport at city-wide scale. In the VivaMK project, Vivacity Labs are  giving Milton Keynes Council the unprecedented ability to understand, model and optimise multi-modal transport networks, while also providing citizens with next-generation journey planning data, such as the quickest route to an empty car parking space.

To achieve this, Vivacity Labs are installing c.425 traffic sensors across c.850 carriageways, and c.1,800 parking sensors across c.12,500 parking spaces along with the cloud based infrastructure, with code using the latest in machine learning to remove privacy concerns, minimise data costs delivering data to bespoke front ends to display and serve the data to clients in standard web browsers. Vivacity Labs are also using the latest in AI research to provide accurate predictive forecasts of the transport network giving 15 minutes, 30 minutes, hourly and day ahead forecasts.

Key challenges and how were these overcome:

  • Challenge 1: Obtaining full internal buy-in from multiple teams in the city to deploy cutting-edge artificial intelligence at scale for the first time. Pre-programme with MK:Smart to build up confidence in the underlying technology. Sequenced internal signoff process, ensuring installations could commence while continuing with discussion around commercials and benefits, allowing short timescales to be achieved. Extensive workshops and demonstrations of progress to bring all teams on the journey.
  • Challenge 2: Growth of both company delivering project & manufacturing / installation capability, from zero to 2,000 sensors over 6 months without compromising quality. Co-located manufacturing with project installation site to maximise speed of iteration; close relationships built between manufacturing team and council operational teams to ensure rapid feedback loops; front-loading of investment (in this case by private investors) ensured that resource was in place to enable rapid growth.

Key success factors:

  • Installation of 425 traffic sensors across major junctions across Milton Keynes.
  • Installation of 1800 parking sensors across paid council operated parking spaces in central Milton Keynes.
  • Deployment of traffic system software pipeline able to serve historical classified vehicle counts and city wide busyness across Milton Keynes to clients including Milton Keynes Council.
  • Deployment of parking system software pipeline able to serve individual bay occupancy status alongside location, costs and bay type (disabled, electric, car sharing) across 12,500 spaces in central Milton Keynes to clients including Milton Keynes council and citizens.
  • Creation of citizen app, MotionMap, to serve parking occupancy status, location, cost, attribute to citizens in Milton Keynes.
  • Creation using latest AI research to provide accurate forecasting of the transport network.

Key learnings:

  • Deploying innovative capabilities with small companies requires flexibility on cashflows & financing
  • Early engagement with teams across council required to enable rapid implementation
  • City-wide transport data adds value to a wide range of teams & users, not just the highways management department

Organisation: Milton Keynes Council and Vivacity Labs

How long has the project been operational: The project commenced in April 2017 and is funded until March 2018.

How long the project took from concept approval to implementation: The InnovateUK funding was approved in March 2017, with the traffic sensor installation complete 8 months later, and parking sensor installation due to be complete 12 months later, by March 2018.

Funding: The system is being funded through £1.7m from Innovate UK sponsored by the Department for Business and through a further £1.3m of private investment raised by Vivacity Labs

Pictures courtesy of: Milton Keynes Council and Vivacity Labs

Share This