Westminster City Council × Veolia
Mobile Capacity Mapping — 2024
Westminster doesn’t have a coverage problem — it has a capacity problem. A million visitors a day try to use the same finite radio spectrum across an eight-square-mile borough. In February 2024, Inakalum partnered with Veolia, Westminster’s waste collection contractor, to place battery-powered NetworkUX kits in their bin lorries. In one week, the trucks captured over 1.2 million datapoints across every Westminster street — the data that became the borough’s evidence base for everything that followed.
- Borough
- City of Westminster, London
- Survey duration
- 1 week (February 2024)
- Datapoints captured
- 1,200,000+ in 8 sq mi
- Networks measured
- Vodafone · EE · O2 · Three
- Fleet partner
- Veolia (Westminster’s waste collection contractor)
Inakalum CEO Philip Maguire and Westminster Head of Smart City David Wilkins.
Filmed in Covent Garden in 2024 — a live walkthrough of the capacity problem, the methodology, the data, and the outcomes the council has been able to deliver from it. Roughly eleven minutes.
“We have 100% coverage. That tells me nothing.”
Two years before the survey, Inakalum CEO Philip Maguire met Westminster’s Head of Smart City David Wilkins to pitch a coverage-mapping engagement. David’s response set the entire project in motion: “Coverage doesn’t tell me anything — we have 100% coverage in the borough. What we have is a capacity problem. If you can find a way to measure that, then we’ve got something to talk about.”
Can you reach a cell at all?
The classic mobile-coverage question: do you have any bars of signal on your phone in this location, on this operator? Westminster’s answer is essentially yes, everywhere. The borough is one of the best-covered places in Europe.
Can you actually use the service?
A million visitors a day in eight square miles. All of them, simultaneously, trying to use the same finite radio spectrum. The phone shows three bars. The apps still time out. That’s the capacity question — and operator coverage maps don’t answer it.
“There are times in places like Westminster where you have a signal, but you’re still unable to use your phone. And that’s because we have around a million visitors a day — a million people trying to use the finite capacity on the mobile networks.” — David Wilkins, Head of Smart City, Westminster City Council
The same spot. The same three bars of signal. Two different networks.
Philip Maguire ran the same set of mobile apps in Covent Garden twice on the same day — at mid-morning when the networks were quiet, and again at 5:30pm when the evening crowd had arrived. The signal indicator on the phone never changed. The apps’ behaviour did.
- YouTube: Loaded fine.
- BBC News (video): Streamed cleanly.
- Google Maps: Painted the map instantly. Blue dot at the right spot.
- YouTube: Spinning wheel. No video would load.
- BBC News: “No network connection.”
- Google Maps: Blue dot but couldn’t paint the map of the area.
“There’s three bars coverage. It should be working fine. It’s giving me a message saying that there’s no network connection. That’s classic capacity problems and how they manifest for users.” — Philip Maguire, CEO, Inakalum
Eight square miles. One week of bin wagons.
Inakalum partnered with Veolia, Westminster’s waste collection contractor, and placed battery-powered NetworkUX kits in the Veolia lorries that already drive every Westminster street as part of their normal weekly routes. Because the kits are portable, they were swapped between trucks each day to target whichever routes covered the most ground that day — ensuring the whole borough was reached inside a single week. Over those seven days the kits continuously uploaded and downloaded files to a cloud server, measuring the actual throughput each of the four UK networks was delivering at every metre along the route — while the trucks did the job they were doing anyway.
The result: 1.2 million+ geolocated, timestamped datapoints across an eight-square-mile borough — the largest dataset NetworkUX had ever captured at the time, and the project that proved the bin-wagon model that has since scaled to Manchester and the Tees Valley.
The same streets. Twelve hours apart.
NetworkUX’s RAG Polygon tool visualises the 1.2 million datapoints across all four operators, with three-hour time-of-day filtering. The two maps below show the same survey routes in central Westminster — one filtered to a 6am window, the other to 6pm. Green is 4+ Mbps (every app works). Orange and red are where capacity has run out.
The transformation between the two views is the capacity story. Coverage didn’t change. Signal strength didn’t change. What changed was the load, and the visible collapse of usable throughput at exactly the times when residents, workers and a million visitors most needed the network.
What the council actually did with the data.
In a sector where most measurement work ends at the report, Westminster turned the Inakalum dataset into a portfolio of policy, infrastructure and operator-engagement outcomes that are still compounding.
DSIT funding secured for smart lampposts
Westminster used the Inakalum data as the evidence base for a successful funding application to the Department for Science, Innovation and Technology (DSIT). The award is being used to test smart lampposts that are physically and electrically capable of carrying operator small-cell equipment.
50 additional small cells from operator investment
By bringing structured, multi-operator evidence to mobile-network operators, the council secured commitments for roughly 50 additional small-cell deployments in the capacity not-spots the survey had identified. The conversation shifted from “please do something” to “here is where, and here is the data.”
Neutral-host policy for Oxford Street
The data revealed that Oxford Street has capacity not-spots on all four UK networks simultaneously. Combined with the high demand on Oxford Street lamp columns for sensors, cameras, hanging baskets and Christmas lights, the council is now exploring a neutral-host model for shared small-cell infrastructure across all four operators in one deployment.
Faster, more constructive operator conversations
With granular per-operator capacity data in hand, Westminster moved from advocacy to collaboration. The conversation with mobile-network operators is now anchored in evidence, and decisions about where to invest are reached in weeks rather than months.
“The bottom line is, now that you have data, conversations are moving much quicker, much faster. There’s actually actions happening where you’re working towards resolving the problems with capacity in the borough at a much faster rate than you were before you had the data.” — Philip Maguire, CEO, Inakalum
“The conversation with mobile-network operators has moved on a lot more. We know where the problem areas are. We can be constructive and collaborative about how we might be able to help them — our aim is to improve the user and visitor experience of people coming to Westminster.” — David Wilkins, Head of Smart City, Westminster City Council
Every street, every measurement, every operator.
The RAG polygons are summaries of the raw measurements underneath. When the council team needs to drill into a specific street or time window, the NetworkUX platform shows every individual datapoint — geolocated, timestamped, attributed to its operator, and colour-coded by throughput.
Four takeaways for boroughs, BIDs and metropolitan councils.
Coverage is the wrong question for cities
In dense urban areas with mature 4G/5G build-out, almost everywhere has signal. The question that matters is whether the radio resource can carry the demand at the times it’s needed. That’s a capacity measurement, and operator coverage maps don’t produce it.
Time-of-day analysis is the differentiator
Westminster’s 6am map and 6pm map are the same streets. A single static coverage view would have completely missed the story. Three-hour time-of-day filtering on the RAG Polygon is what made the capacity problem visible — and the case for investment defensible.
The bin-wagon model travels
Westminster proved that fleet-equipped surveying produces a borough-scale dataset in a single week with no extra fieldwork. Manchester repeated the result for a major city in four weeks. Tees Valley scaled it to five councils across six months. The methodology is reproducible at any geography.
The data is the policy lever
Westminster used the same dataset to win DSIT funding for smart lampposts, to unlock 50 small cells of operator investment, and to underwrite a neutral-host model for Oxford Street. One survey, multiple compounding policy and infrastructure outcomes — that’s what evidence-led council work looks like.
The bin-wagon model can map your borough in a week.
Westminster’s engagement took a single week of Veolia’s normal routes to produce the dataset that has shaped the borough’s digital connectivity policy since. The same pattern works for any council with a waste-collection fleet and a capacity question.