We know running a red light is dangerous, but the truth is we don’t have the data to know how dangerous. Let’s say you’re taking a safe systems approach to redesigning a junction: you’ve done all you can to slow vehicle speeds and separate those most at risk. But it still needs a level of compliance to function. Do you need enforcement to maintain the integrity of the protective environment that you’ve built?

Normally it’s the Police that decide where enforcement cameras go, based on prosecutions and revenue. It’s a useful and effective “sinner tax”. How much better would it be if you could measure compliance as one factor of the safety of a junction, and get hard data on if it’s a problem? And then use that to pull the right lever to save lives.

Working with Transport for Greater Manchester, VivaCity is exploring how innovations in AI-powered Computer Vision sensors can bridge this data gap. This type of technology can provide data to let you assess what the right intervention is or measure if your safe systems approach is working. Are pedestrians and cyclists using the infrastructure you expect? Are there dangerous speeds? What are the most common “near miss” road user interactions?

By layering red light running on top of these features, you then have context on if unsafe behaviour was related to infrastructure compliance; a safe systems approach as opposed to a limited enforcement question. This means the right areas can be targeted with the right safety levers, giving you better, more effective and efficient safety outcomes.

At NRSC 2026, we look forward to sharing an early look at this innovative diagnostic capability. We will explore how contextual, AI-powered safety data empowers transport authorities to pull the right levers to effectively and efficiently save lives.

Cleo Weeden, Strategic Partnership Lead, VivaCity