Langeland: 220 sensors, one vehicle less

The round before

Every cycle, an employee drove a 176 km round trip across the island to check containers at 50 recycling stations. It took a day and a half and produced a paper report listing fill levels. A driver then used that report to plan which containers to empty.

Two problems, beyond the obvious cost. The process was cumbersome, and the assessment varied depending on who made it. A container one inspector called nearly full was, to another, comfortably fine for another week. The operations manager's summary was that far too much time was being spent checking the municipality's environmental stations.

What was deployed

The utility approached it systematically, starting with about ten sensors and scaling to 220 — one for every container. The island now has around 30 collection points covering, among other areas, all holiday homes with 2,300 households. Each point holds six or seven containers and handles ten waste fractions in total. To make collection more efficient, glass and metal are collected as a single fraction.

Alongside the sensors, the utility took the transport and logistics software that retrieves data from every container via the cloud, generates collection routes automatically, and gives drivers the service route and current fill levels through an app.

Fill levels update automatically every six hours. A container is normally added to the service route at 80–90% full. Every Wednesday, an additional check is run for containers that might become critical over the weekend in holiday-home areas — a small local rule that prevents the failure mode most likely to generate complaints.

The results

• The 176 km manual inspection round was eliminated.
• Combined with a move to consolidated collection points, the utility reduced its fleet by one waste collection truck.
• Overflowing containers in the holiday-home areas ceased to be a problem.
• Reliability: one defective sensor out of 220 in the first four months, and no battery replacements across the year.

The operations manager's stated goal, delivered with some satisfaction, is no letters to the editor of the local newspaper about missed collections. It still holds.

Connectivity on difficult terrain

Langeland is a beautiful island of rolling landscapes, and that initially caused problems transferring data from sensors distributed across it. NB-IoT was the technology that resolved it, providing consistent data transfer about container fill levels. For any UK authority servicing rural bring sites, upland car parks or coastal locations, this is the relevant precedent: NB-IoT reaches places ordinary connectivity does not, because it was designed to.

Expanding is a ten-minute job

The utility runs a subscription covering both the sensors and the software. When a sensor needs installing or replacing, the operations manager frequently does it himself while already out on the road — ten minutes to install a new fill-level sensor and set it up in the software.

That figure matters more than it appears. A deployment where adding a container requires a supplier visit does not grow; one where a supervisor can do it from the cab does. Langeland went from ten sensors to 220 on that basis.

What they were planning next

Two things, both worth noting because they mark the maturity curve of a deployment. Moving to the next generation of radar fill-level sensors for greater precision on cardboard, paper and plastic — the materials that defeat older measurement technology. And beginning to use emptying forecasts rather than reacting to current fill levels alone.

The operations manager recommends other waste handling companies install fill-level sensors, on the grounds that it creates obvious opportunities to optimise operations and that the more rational effort also makes good sense for the employees.