Case study · Denmark
Vejle: rounds that rebuild themselves at 05:00
A municipality-wide rollout across 2,000km of roads — and the clearest published account of why radar measurement matters.
Vejle Municipality fitted fill-level sensors to 547 of its 875 public waste bins, spread along 2,000 km of roads. An optimised collection round is now calculated automatically at 05:00 each morning. Average collection intervals moved from once or twice a week to once a fortnight, and staffing on the task fell from two full-time employees to one and a half.
From trial route to council decision
Vejle allocated funding for a pilot on a single trial route, with the objective of replacing estimated-frequency waste collection with data-driven planning. The test was intended to establish three things: whether collection could be optimised, whether service to citizens improved, and whether waste management became more environmentally sound.
It demonstrated clear optimisation potential and a positive business case, and the fill-level sensors were subsequently rolled out across the municipality following a council decision. This sequence — funded pilot, measured business case, political decision, phased rollout — is the one we recommend to UK authorities, for the same reasons.
The deployment
547 of the municipality's 875 public bins now carry radar sensors mounted on the inside of the lid. The remaining 328 sit in central locations in Vejle and the municipality's main towns, where they are emptied daily by staff as part of other routines — there is no wasted journey to eliminate, so there is nothing for a sensor to save.
That distinction matters, and it is a point of discipline worth borrowing: instrument the containers where a vehicle travels specifically to reach them, not the ones already passed every day.
Sensors flag a bin as ready for emptying at 80% full and report six times a day over NB-IoT. The platform collects the readings and generates optimised rounds online for the municipality's internal contractor, the Materielgården, which carries out the collections.
The results
- Collection intervals moved from once or twice a week to an average of once a fortnight.
- Staffing on bin emptying fell from two full-time employees to one and a half, with scope for further reduction through ongoing optimisation.
- The freed capacity was redeployed — with the same staff, the team also now collects waste along roads and at road bridges.
- Programme targets: 30% reduced travel, a 40% economic gain and a reduction of 5.5 tonnes of CO₂ per year.
Previously we would drive out to the bins once or twice a week, but now we manage to empty them on average every fortnight.Operations assistant, Roads & Parks — Vejle Municipality
Why they changed sensor supplier
This is the most useful part of the Vejle story for anyone writing a specification. The original 2021 trial used non-radar fill-level sensors and produced error reports on up to 30% of readings. That level of false reporting is fatal: once operations staff have been sent to bins that turn out to be half empty, they stop acting on the data, and the project dies regardless of what the dashboard shows.
Vejle switched supplier specifically to obtain radar-based measurement, and now experiences very few error reports — describing the radar sensors as performing clearly better than the goal set for the test period. Signal coverage over NB-IoT was also reported as very good, with fill levels updating six times a day.
If you take one line from this case study into a procurement specification, make it radar.
The forecasting layer
Round planning uses an AI forecasting tool that improves over time at predicting when each individual container will next need emptying, combining current sensor measurements, historical fill patterns, drivers' reports and other data sources. The practical benefit is consolidation: a container approaching threshold can be collected while a vehicle is already in the area, rather than triggering its own journey days later.
What made it work
Vejle's team are unusually candid that the hard part was not the technology. Implementation required discipline and fixed routines in registering every bin with its type and GPS location in the municipality's database, and one person responsible for coordinating across the departments involved and for keeping that location data correct.
The second requirement was driver buy-in. The operations assistant responsible for the rollout makes the point directly: it is important that drivers support the digital solution, and they do, because it is easy to use — including for substitutes. A system that only the regular driver can operate fails the first time someone is off sick.
The political dividend
An outcome the authority did not initially anticipate: the ability to answer elected members quickly when residents raise questions about bins. The head of Roads & Parks notes that politicians like being able to show that unnecessary journeys are not being made, because it demonstrates that the climate has been considered. For a UK council with a declared climate emergency, that is a directly transferable argument.
Deployment at a glance
- Location
- Vejle Municipality, Denmark
- Estate
- 875 public waste bins across 2,000 km of roads
- Instrumented
- 547 bins
- Not instrumented
- 328 central bins already emptied daily as part of other routines
- Technology
- Radar sensors mounted inside the bin lid, NB-IoT
- Reporting
- 6 readings per day
- Threshold
- Flagged for emptying at 80% full
- Round generation
- Automatic, calculated at 05:00 daily
- Collections by
- Materielgården, the municipality's internal contractor
- Interval before
- Once or twice a week
- Interval after
- Once a fortnight on average
- Staffing
- 2.0 → 1.5 full-time employees
- Targets
- 30% less travel, 40% economic gain, 5.5 t CO₂/yr
- Earlier trial
- 2021, non-radar sensors, up to 30% error reports — supplier changed
- Fully implemented
- 2023
These figures are as published by the operating authority. They document a deployment of BrainyBins® by Maacks ApS — the platform Britannia IoT Solutions supplies in the United Kingdom — and are presented as evidence for the technology rather than as a Britannia IoT Solutions client engagement. Outcomes in any given operation depend on current collection frequency, bin density and round geography — which is what a pilot measures.
Questions
Reading the Vejle deployment
Why did they only instrument 547 of 875 bins?
Because the other 328 are in central locations already emptied daily by staff as part of other routines. There is no journey to avoid at those bins, so a sensor would produce information nobody would act on. It is a good discipline to copy: instrument the containers a vehicle travels specifically to reach, and leave the ones already passed every day alone. It also keeps the subscription cost proportionate to the saving.
What does the 30% error rate on the earlier trial tell us?
That measurement technology is the whole ball game. Vejle's 2021 trial used non-radar sensors and produced error reports on up to 30% of readings, which is enough to destroy operational confidence entirely — a supervisor sent to two empty bins stops trusting the third flag. The authority changed supplier to obtain radar measurement and now reports very few faults. When writing a specification, the word to insist on is radar, not simply “fill-level sensors”.
How does a fortnightly interval not cause overflow?
Because it is an average outcome, not an imposed frequency. No bin is on a fortnightly schedule; each is collected when it reaches 80% full, and the fortnightly figure is simply what that works out to across the estate. A busy bin is still emptied often. What disappeared is the routine journey to bins that were nowhere near full — which is also why service to residents did not deteriorate.
Did Vejle reduce staff?
From 2.0 to 1.5 full-time employees on bin emptying, with the freed capacity redeployed rather than removed — the same team now also collects waste along roads and at road bridges, work that had not previously been getting done. The authority expects further optimisation to reduce the emptying requirement more. It is worth being straightforward with crews about this from the outset, because they will identify the implication immediately.
Next step
See what your own bin data would say
A 90-day pilot on one round tells you exactly how much of your current collection effort is going to bins that were not full. No commitment beyond the pilot, and you keep the data.