Research project
McMatcher: Route detection and Proximity Analysis
- Start date: 30 November 2024
- End date: 30 November 2025
- Funder: EPSRC
- Value: £60,000
- Partners and collaborators: Mott McDonald
- Primary investigator: Professor Susan Grant-Muller
- Co-investigators: Dr Andrew Mark Tomlinson, Yanis Boussad, Dr Yuanxuan Yang
- External co-investigators: Katy-Anne Moseley (Civil Engineering, Leeds)
The project team at the Institute for Transport Studies at the University of Leeds has developed the innovative new “McMatcher” technology, which can sense individuals moving past strategically placed devices by detecting commonly carried smart devices (phone, smartwatch etc).
The McMatcher technology comprises 1) a battery-powered sensor with data storage and 2) an advanced data processing methodology to extract insights for location, trajectory and proximity monitoring.
Using the data collected we can construct a probabilistic movement pattern for each person passing through an area. Without GPS tracking, McMatcher identifies movements between levels, for example, when taking a lift or stairs. The main novelty is a methodology for device reidentification and individual-tracking, achieved without breaching device-based privacy-protecting measures (e.g MAC address randomisation). In this project we will benchmark McMatcher as a novel data collection technology against manual surveys (completed by Motts McDonald) and CCTV-based image processing methods at a train station. Motts Macdonald specialise in modelling pedestrian flows on a commercial basis for governmental agencies and industry. However, manual surveys are costly; inaccurate in crowded settings, rely on subjective classifications, are limited to points that are safe for human observers/enumerators to stand and are only carried out over limited hours/time periods. Alternative CCTV-based image processing methods by contrast, have public acceptability issues, are costly, inaccurate and limited to particular ranges of vision. Other signal-based approaches can capture total footfall but not detailed individual routes through a site or, importantly, movement between floors/levels. McMatcher could provide a solution to a challenging weakness in pedestrian flow modelling i.e. difficulty capturing fine-grain data on individuals’ choices and movements through transport interchanges (and other buildings).
Insight from McMatcher, could underpin a wide range of policy and delivery improvements in transport services and infrastructure provision. This will in turn, help deliver efficiency and safety benefits and reduced environmental impacts, thereby promoting active travel and demand for rail and bus services.
Impact
We envisage that McMatcher will give more accurate and detailed pedestrian flow data. This could support better planning of transport terminus and travel service demand, with broader environmental (carbon) benefits.
Furthermore, increasing our understanding of individual variations in the routes and time taken through different places, such as shopping centres, hospitals, places of education could support more equitable planning of services, for example for those with mobility challenges.
Finally, with the ability to understand demand and movements there is potential for the technology to be used in improving infrastructure security and resilience planning, aligning with the priorities of UK public bodies (DfT, UKHSA, RSSB, DSTL) and more broadly the UK Resilience Framework, to build resilience and preparedness for threats.
Publications and outputs
https://eprints.whiterose.ac.uk/213511/1/McMatcher_final.pdf