As part of Greater Manchester’s commitment to Vision Zero, this project pilots AI-enabled adaptive roadside signage designed to influence real-time road user behaviour. Using camera-based detection and edge AI processing, the system identifies contextual risk factors — including speed profile, vehicle type, pedestrian and cyclist presence, time of day, and traffic conditions — and delivers tailored safety messages via dynamic message signs.

Unlike traditional static or trigger-based signs, the intervention adapts messaging to the immediate road environment. For example, where vulnerable road users are detected, signage prioritises pedestrian or cyclist awareness messaging; where speed non-compliance is identified, behaviour-specific prompts are displayed. The aim is to increase situational awareness, encourage micro-behavioural adjustments, and reinforce shared responsibility within the Safe System framework.

The intervention is positioned as a behavioural change within a wider engineering and enforcement strategy. Importantly, it moves beyond generic warnings toward contextual, data-informed communication at the point of risk.

Evaluation is embedded from inception. A controlled before-and-after methodology is being applied, incorporating speed and compliance data, near-miss proxy indicators, and comparative site analysis. Behavioural response patterns to different message types are also being examined to refine effectiveness. Early findings will explore whether adaptive, context-specific messaging produces greater compliance improvements than standard reactive signage.

This presentation will outline the technical model, governance considerations (including data protection), partnership working, and initial evaluation insights.

Adam Meadowcroft, Thriving Places: Infrastructure and Network Manager, Stockport Council

Biog to follow.