Client
A UK strategic road network operator responsible for incident response and traveller information across a regional motorway and trunk road network
Sector
Transport & Infrastructure
Engagement
Live incident coordination platform linking traffic officer status, police liaison updates, contractor clearance sign-off and roadside variable message signs into one shared operational view - multi-quarter programme.
What the client needed
Reopening a carriageway after an incident depends on several separate confirmations lining up: traffic officers confirming the carriageway is physically clear, the police confirming their investigation or recovery work has finished, any contractor confirming barriers and debris are cleared, and the regional control room updating the variable message signs and public traffic feed to say so. At our client's control room, those confirmations arrived by radio call, were logged by hand on a whiteboard, and were then manually keyed into the sign control system by whichever operator was free. Nothing forced the last step to happen quickly once the first three were done - a carriageway could sit fully clear and driveable for the better part of an hour while signs upstream still showed a closure, because the confirmation had reached the control room by radio but hadn't yet been typed into the system that actually changed what drivers saw. The operator's own incident logs showed the average gap between the last clearance radio call and the signs updating was 38 minutes, and on longer incidents that gap alone was adding meaningfully to lost network capacity that had nothing to do with the incident itself.
How we worked
- Built a live incident board that captured traffic officer, police and contractor clearance status directly from radio-linked field updates, replacing the whiteboard as the single record of what had actually been confirmed.
- Connected the incident board directly to the variable message sign control system, so a fully confirmed clearance could trigger a sign update without a manual re-keying step in between.
- Kept a human confirmation step before any sign change went live, since a false or premature "all clear" reaching drivers is a worse outcome than a slower one - the platform removed re-typing, not judgement.
- Integrated a shared status view for the police liaison desk, so their confirmation reached the same board directly rather than via a separate call into the control room.
- Ran the platform in shadow mode against live incidents for six weeks, comparing the board's timings against the existing radio-and-whiteboard process before it took over sign updates for real.
- Trained control room operators and the police liaison desk together, since the timing gain only holds if both sides are updating the same board rather than one side quietly reverting to a phone call.
Measured results
All figures verified with the client. Specific incident, location and personnel detail withheld in line with our standard confidentiality terms.
- Average gap between a carriageway being confirmed clear and the roadside signs and public feed reflecting it fell from 38 minutes to under six.
- Police liaison confirmations now land on the same shared board as traffic officer and contractor status, rather than depending on a separate call into the control room.
- The shadow-mode period surfaced a recurring gap where contractor clearance was being confirmed on site but wasn't reliably reaching the control room at all under the old process, since corrected in the board's alerting.
- Control room operators report materially fewer instances of signs showing a closure the road no longer has, a complaint that previously reached the operator through its own traveller feedback channel.
- The control room is now extending the same board to cover planned roadworks handovers, beyond the incident-response scope the programme was built for.
"The road was clear for most of that 38 minutes - the signs just hadn't caught up. Drivers don't experience our radio log, they experience the sign, so that was always the gap that mattered most and it was the one we'd stopped noticing because it wasn't anybody's fault, it was just how long re-typing took."
Working on something similar?
If this engagement looks like the kind of problem you are facing, we would be glad to compare notes by email.
Context and constraints
A regional road control room doesn't control every party whose confirmation it depends on: traffic officers and the control room itself sit with the operator, but the police retain their own command structure for any incident involving an investigation, and contractors doing barrier or debris clearance are typically third parties on their own call-out contracts. That meant the programme was never going to be a single closed system - it had to be a shared board that each party fed into on their own terms, without any one of them giving up their existing command and control arrangements. The brief was explicit about that from the outset: build the fastest possible route from "confirmed clear" to "signs updated" without asking the police liaison desk or contractors to adopt a system that wasn't theirs to adopt.
The sign control integration turned out to be the more constrained piece of engineering, not the incident board itself. The existing variable message sign system had been built for operator-typed input, not for a downstream system triggering it, and its change-control process required a human confirmation step for good reason - an incorrect "all clear" reaching drivers on a live carriageway is a materially worse failure than a slow one. We built the integration to remove the re-typing step specifically, while keeping a deliberate human confirmation click before any sign actually changed, which meant the timing gain came from cutting transcription lag, not from cutting out judgement.
Building a board three separate commanding parties would actually feed
The harder problem wasn't the integration, it was that traffic officers, the police liaison desk and contractors had no shared habit of updating one system, and a new board doesn't earn that habit just by existing. We ran the platform in shadow mode against live incidents for six weeks, logging what the board showed against what the whiteboard and radio log confirmed manually, before it was allowed to drive real sign changes. That shadow period surfaced a gap nobody had previously had reason to notice: contractor clearance was often confirmed verbally on site to whichever traffic officer happened to be nearby, but wasn't reliably making it back to the control room as a distinct confirmation at all - officers assumed someone else had logged it. That gap is now closed directly in the board's own alerting, which prompts explicitly for a missing contractor confirmation rather than assuming silence means done.
Designing for the moment drivers actually experience
A control room's internal timeline and a driver's actual experience of a closure are two different things, and it's easy to optimise the first while leaving the second untouched - which is exactly what had happened before this programme, since the incident itself was usually cleared efficiently and the delay was entirely downstream of that. The board was deliberately built around time-to-sign-update as the headline metric, not time-to-clearance-confirmed, because the second one was never the operator's actual problem. That framing, more than any single integration, is what kept the programme focused on the 38-minute gap rather than on incident response speed that was already reasonably good.
Lessons learned
The first lesson was that a shared board across parties with genuinely separate command structures only earns trust once it's run in shadow against the existing process for long enough to catch the gaps nobody knew they had - six weeks was what it took here, and cutting that period short would have meant discovering the contractor confirmation gap after cutover instead of before it.
The second lesson was that removing a manual step and removing judgement are not the same thing, and conflating them would have been the wrong trade for a system that changes what drivers see on a live carriageway - keeping the human confirmation click was a deliberate design choice, not a leftover of the old process.
The third lesson was that the metric worth optimising is rarely the one an organisation already tracks well - incident clearance speed was already good here, and the real gain was in a downstream step nobody had been measuring as its own thing until the whiteboard was replaced with something that could time it.
If your organisation has a downstream public-facing step that lags well-run operational work upstream of it, we would be glad to discuss what a programme like this might look like for you. Email sales@halfteck.com.