Browse
Log in / Register
Share
March 5, 2025 – Wiltshire Council has pinpointed and fixed a dangerous junction outside Westbury by analyzing three years of connected vehicle data and onboard camera imagery, bypassing the traditional wait for collisions to mount.
Residents on Trowbridge Road had reported that drivers frequently failed to yield to traffic on Bratton Road, a priority through route. Instead of simply dispatching staff for spot checks, the council turned to anonymized telematics data from Compass IoT, a provider that collects information directly from vehicles already on UK roads.
The data revealed granular, meter-by-meter metrics: speeds, g-forces, and braking patterns across 2024, 2025, and 2026. “Using this type of data allows a different insight into understanding the issues motorists encounter on the network,” said Gareth Rogers, Traffic Engineering Manager at Wiltshire Council.
On the priority route (Bratton Road), average free-flow speeds ranged from 35.8 to 37.9 mph. But g-force readings showed a sharp increase in harsh braking: from 0% of westbound vehicles in 2024–2025 to 16% this year, and 3% for eastbound vehicles. On Trowbridge Road, where drivers must give way, northbound speeds varied widely—from 19 to 43 mph—compared to more consistent patterns in prior years. Harsh braking on that approach jumped from 5% (2024) to 21% (2025) northbound, and from 3% to 22% southbound.
The data pointed to a worsening conflict point, but the cause remained unclear until the council paired the telematics with street-level imagery from Compass IoT’s connected-vehicle fleet. Comparing 2024 Google Maps imagery against July 2026 Compass IoT images, engineers spotted the culprit: roadside hedges had grown so thick that they blocked drivers’ view of oncoming traffic exactly at the moment they needed to slow and decide whether to proceed.
“The hedging had grown significantly over the past two years, obscuring the give-way sign and the view of approaching vehicles,” a council spokesperson noted. “This explained why we saw more drivers braking hard at the last second—they simply couldn’t see the threat.”
Wiltshire acted quickly, trimming the vegetation and restoring sightlines. Recent vehicle imagery captured this week shows the approach now clear, with signage visible again. The fix was validated by the same data stream: harsh braking percentages are expected to drop as drivers regain full visibility.
The approach marks a departure from traditional UK road safety audits, which rely on resident complaints, STATS19 collision data, and periodic site visits—methods that require problems to accumulate before action is taken. Connected vehicle data, by contrast, reveals risk signals early, such as erratic braking and speed variations, allowing engineers to intervene before crashes occur.
“This is a textbook example of predictive safety,” said Dr. Emily Hart, a transportation researcher at the University of Southampton (not involved in the project). “Most councils are reactive. Using real-time behavioral data from thousands of vehicles lets them spot hidden hazards that no single site visit could catch.”
The technique is gaining traction in the US as well: state DOTs in Florida and California have piloted similar telematics programs to identify high-risk intersections and work zones. Wiltshire’s use of a three-year historical dataset, however, is notably thorough, offering a before-and-after benchmark that demonstrates the value of longitudinal vehicle data.
Wiltshire Council now plans to extend the data-driven approach to other junctions with known safety concerns. “We don’t need to wait for a crash report to know something’s wrong,” Rogers added. “The data tells us where drivers are struggling, and we can fix it before anyone gets hurt.”
(Picture: Compass IoT, supplied by Wiltshire Council)
