From reactive safety to early risk signals
Urban cycling safety policies typically rely on reported accident data, information that only becomes available after incidents have occurred. In a pilot experiment in Mechelen, researchers explored whether continuously collected traffic data, gathered with inexpensive Telraam sensors could offer earlier insights into where and when cyclist‑car interactions may pose higher risks.
The goal was not to predict accidents, but to test whether simple, scalable indicators derived from these sensors could surface meaningful patterns worth further investigation.
Eight streets, one composite indicator
The pilot focused on eight street segments representing different urban contexts, ranging from quiet residential streets to busier connectors with school‑ and pedestrian‑related traffic. Using Telraam’s existing sensors, the team collected hourly data on cyclist counts, motor vehicle counts and vehicle speeds over periods of up to one year.
From this data, a composite risk score was constructed by combining:
- cyclist volume
- motor vehicle volume
- the share of cars exceeding 40 km/h
The underlying assumption is intuitive: traffic risk is shaped not just by speed, but by how many road users interact at a given time.
Patterns that invite closer attention
Across the eight locations, the composite scores varied considerably. Temporal patterns, by hour of day and day of week, differed between streets in ways that are not immediately obvious from general knowledge of those locations.
These variations suggest that even a simple indicator may help surface micro‑level patterns that could inform targeted, low‑impact interventions, such as:
- timing of speed signage
- school zone activation windows
- prioritisation at shared junctions
While the pilot cannot support policy decisions on its own, it highlights promising patterns worth examining.
Correlation, not causation – and why that matters
The team also explored correlations between the composite scores and historical accident data from Statbel (2017–2024, filtered for cyclist–motor vehicle incidents). A pattern‑level correlation was observed at the hourly scale.
However, the study explicitly avoids causal claims. Street geometry, intersection design, visibility and signage, all factors not captured by the sensors, are likely important contributors to cycling safety and remain outside the scope of this pilot.
This transparency reflects CitCom’s emphasis on responsible experimentation and careful interpretation of early‑stage results.
A scalable direction for cities and companies
What the experiment in Mechelen demonstrates is that the research question itself is worth pursuing further. Telraam has disseminated these results on a bike event in Belgium and was well received by The Fietsersbond (Cyclists' Union). With larger datasets, additional contextual variables and testing across a wider range of cities, a validated composite risk indicator could eventually become a cost‑effective, scalable tool for proactive cycling safety prioritisation. Telraam is already updating its backend data structure to better collect the characteristics of the streets in order to make this evolution possible.
Such an approach could also be used in traffic simulations, allowing cities to assess how new circulation or mobility measures might influence risk patterns before they are implemented and what the impact is of changes in infrastructure.
For Telraam, the Mechelen results confirm a clear path from pilot to service offering. Cities get a fairly simple, scalable and evidence-based safety tool at a fraction of the cost of traditional monitoring. That's the step that moves a good research idea into day-to-day municipal practice.
- Kris Vanherle, CEO Telraam
Experimenting safely with CitCom.ai
This exploratory study was conducted together with Telraam within the CitCom.ai TEF in combination with the Smart Infrastructure Facilities (SIF) programme, using open data from the Telraam API and Statbel. It illustrates how CitCom enables cities, companies and researchers to test ideas quickly, using real‑world data, before committing to large‑scale investments or policy changes.
Interested in exploring similar low‑cost mobility or safety experiments in your city?
Have a look at our other services: https://citcomtef.eu/services
