Client: Fyma (computer-vision traffic analytics), for the London Legacy Development Corporation
Location: Queen Elizabeth Olympic Park, London, UK
Year: 2021
Sector: Computer Vision · Mobility KPIs · Product R&D
Turning live CCTV computer-vision data into planning-grade mobility KPIs for London’s Olympic Park — a genuine two-way R&D collaboration that fed improvements back into the sensor company’s own product.
Planning authorities usually build mobility KPIs from expensive, one-off surveys. Fyma’s computer-vision platform, reading existing CCTV feeds without storing video, offered something different: continuous, GDPR-compliant traffic data — if someone could turn it into planning-relevant metrics.
A research collaboration between SPIN Unit and Fyma, delivered for the London Legacy Development Corporation, the planning authority for Queen Elizabeth Olympic Park, running through two phases in summer and winter 2021.
Designed a suite of mobility indexes computable directly from Fyma’s ~30-camera network — traffic density, modal share, modal consistency, amenity diversity and popularity — applied across five clusters around the park (Trampery, Stratford International, Westfield Stratford City, the Aquatics Centre, the Stadium).
Delivered a summer analysis and report (August 2021) followed by a winter update (December 2021), alongside a set of concrete suggestions for improving Fyma’s own sensing platform for urban-planning use cases.
- Vehicles accounted for 67.7% of all observed traffic across the study area, reaching up to 90% car modal share on weekends in the Trampery area — already at capacity during peak hours.
- Westfield Stratford City was the one clear exception, where walking consistently outranked driving — correlating directly with that area’s higher mix of necessary and optional amenities.
- Despite high-quality cycling infrastructure, bicycle modal share peaked at only around 8% — the report specifically recommended better east-west cycle connectivity to link the park’s internal network to central London routes.
- Used the same granular camera data to study traffic on a school access road, finding measurable shifts in mode share tied to the school calendar — a concrete road-safety planning input.
- The engagement fed back into Fyma’s own product: SPIN Unit proposed specific technical improvements (like treating pedestrian crossings as measurable micro-zones) rather than just consuming the platform’s output.
Gave the London Legacy Development Corporation planning-grade mobility KPIs derived from existing camera infrastructure — at a fraction of the cost of traditional survey-based methods — while directly improving the underlying sensing platform.