How does the tracking with the CleanHub software work?

Tl;Dr: All our collection partners use our track and trace technology to record the collected volumes. We track the waste from collection to final recovery, where the material goes through a laboratory. One element of the lab tests is a moisture test. All waste always has a bit of water in it. We deduct the moisture from the delivered waste to have the actual dry value, and no one pays for water disposal. Once we know the final dry-waste volume, our app splits it among all brands that support the specific project that collected the waste.

How does the Cleanhub system work?

Our collection partners and the collection points below a specific collection partner are part of our digital ecosystem. Every co-processing partner is part of that ecosystem as well. And finally, all these partners are connected to your specific impact report through our platform.

This all sounds very confusing, so maybe you would like to watch this video to get a better understanding of how it works:

And for an even more detailed overview - here is how the different elements look like:

IMG_2366 (1)

Disclaimer before you start reading: We are aware that a lot of this sounds like we are another of those evil tech companies that are pulling data left and right to make a lot of money.
Two notes on that:

    1. We’re not collecting personal data, only processing data. 

  1. We have deep-rooted trust relationships with our partners; they know what data we’re collecting. There is no reason not to trust them, as we had zero incidents in more than 16 months in operations.

We’re proud of the achievement of having collected more than 500t, including all data points. But this is just the beginning. There are 11 million tons of plastic entering our oceans annually. We want to build a system that works at scale. 

If we operate at scale, there will be hundreds of collection partners on our platform. Collection partners that will benefit from a system that converts negative-value plastic into hard cash. 

There will be bad apples in that crowd. We want to protect all upright collection partners, brand partners, and our own credibility - and catch fraudulent behavior when it happens. The only way to do that without spending massive amounts on auditing services (which would dilute money away from true impact) is through data and ever-increasing pattern recognition in that data.


Let’s start at the beginning of the collection

The journey starts differently depending on the collection method (check out this article to learn more about how we collect waste). In some cases, we go all the way to the household with our tracking. In that case, the collection crew scans a QR code on the door that opens our app and registers that they have now checked in to a specific location to collect waste. 

Once the homeowner hands over the waste, the collectors put it on a small scale. The collection crew takes a picture of the waste bag on the scale in that location. Along with the bag’s image, we resolved the geo-location to make sure the waste was collected where it was supposed to be collected. The collection crew also registers the weight displayed on the scale. We can cross-check these values. 

While the image does not directly show what material is in the bag, we’re getting a good understanding of which weight distribution is normal. The content of the waste bag is not too crucial at this stage because it can still be mixed household waste that will be sorted later.


The collection crew continues the route until they’ve been to all households. Our app aggregates the sum of collected waste. Once they complete the route, the waste goes into a so-called MRF - a Material Recovery Facility. At this stage, all the waste goes on a bigger scale. Usually, the truck that carries the waste is weighed. That way, we get a second data point that we can cross-compare to the collected volumes en route. If there is too much of a deviation, we know that something went wrong. If it is way more, it means that someone is potentially trying to sneak in waste we don’t know the source of. If it’s a lot less, we also know that something went missing on the way.

Whether the waste is sourced from a cleanup, river barrier, or scrap dealer, we always follow that process.

Let’s sort things

Since no collection project only collects multi-layer packaging - the material we’re spending money on - it needs to be sorted. The sorting crews sort out all recyclables so no resources, that could have had a second life, end in co-processing. Also, all non-recyclable plastic waste is sorted out. 


This material is either baled or put into big packs for storage and later transportation. Every bale and every big-pack is individually registered on our app again.

That means we know at every point how much material went into a sorting process and what the output is. Since we also know where the waste was sourced, we are learning over time what standard behavior looks like and what anomalies look like. If there are suspicious data registrations, we can immediately intervene and ask for more information.

For the waste management nerds: we’re monitoring the mass balance.

Time to say goodbye

Once all material is sorted and we have enough volume, it makes sense to put a truck on the street. The crew loads and secures the material and sends it to one of our co-processing partners. On departure, the truck goes over a weighing bridge again, and our collection partners register the net weight of the truck. At this point, we know exactly how much plastic waste left the location (including the geo-location again) and where it is supposed to go. We also know if there is more waste loaded on the truck than the sorting center had in stock - which would again be a sign that there is waste put on the truck that is not supposed to be there. Waste of which we don’t know the source of.

LAL04920 1

On arrival of the waste at the co-processing center, the truck goes over a weighing bridge again. Our co-processing partner registers the weight of the incoming truck. At that moment, we cross-check the volumes booked at the sorting center and those that arrived at the co-processing planet. If there is too much of a difference - the same thing as earlier - something must be wrong.

To the lab and beyond

The engineers at the co-processing location take a sample of every truckload that arrives. This lab result is crucial for the co-processor and us as well. And the co-processor depends on a correct result as much as we do - because they observe three values—the moisture content in the waste, the calorific value, and contamination. 

Moisture content

Moisture is vital for the co-processor because they can’t feed material to the kiln that is too wet and prevents them from reaching the necessary temperatures during incineration. For us, it’s important because we want to issue plastic credits - not water credits. We pay our collection partners based on the dry-waste weight. So if there is a truck of 10t going into co-processing and shows a moisture content of 10%, we pay for a collection of 9t. At the same time, we also only sell 9t - or we only issue 9 plastic credits, to be more precise.

Calorific value

The co-processor cares about calorific value because it helps create the fuel needed to reach the temperatures in the kiln they need in each step. For us, it is yet another data point that helps us prevent fraud. Different plastic types show other profiles of calorific content. While not super precise, massive deviations would again show that it is not the material we wanted to reach the plant.


The lab is looking for specific contaminants in the waste—especially chlorine. Chlorine is part of a particular plastic-type called PVC. Chlorine is problematic for two reasons: the toxic emissions from burning it and chlorine in high concentrations would eventually shut down the plant because it can block the kiln. 

For us - you guessed it - this is once more a data point that can show us that the partners collected waste we didn’t want to be managed.

Finally, we receive a co-processing certificate that guarantees that the waste will be destroyed through co-processing. Once all this documentation is in place, we count the waste as verified and collected.


Why are we going this way?

While we didn’t publish this procedure as a standard yet, we still believe that it is the most comprehensive and inclusive approach from a monitoring and scalability point of view.
Waste management needs to happen locally because many cultural hurdles need to be overcome. 

These communities and their local entrepreneurs don’t need some smart western people to explain the world to them. They need access to capital. We keep investments accessible to even the smallest communal projects. Instead of asking projects to pay us 100k EUR upfront to be registered as a project on our registry, we provide them with the technology to prove that they are doing good work. 

We scanned all the extensive standards that are currently under drafting. None of these standards have a credit-issuing event or describe how a project should monitor its operations. Some of the standards issues credits just for the collection of plastic. What happens to the plastic afterwards is entirely up to the project. Basically, in the worst case, they can dump the plastic on the street and collect it again to receive the credit again and again. So if you evaluate projects, ask them: what happens to the plastic and what defines a credit-issuing event.

We focussed our approach on identifying different types of proof of work. And then tried to make the documentation as easy as possible. No monitoring in waste management can conclude just measuring or taking one single data point out of context. At CleanHub, for every tonne of verified plastic, we receive on average 59 data points. All of these form a bigger picture that tells us what’s going on and whether this is legitimate or not.

By the way, you can monitor all these data points at full transparency on our website:

How are the plastic credits distributed?

At the time of writing, we had five collection partners live, including their supply chains. These five collection partners are “supplying” their volumes to 102 customers. Let’s assume the truck mentioned earlier in the example had the verified weight of these 9t. 

The collection partner that aggregated these volumes will receive the payment for this 9 t. At the same time, all brands that are connected to this project will receive a cut of this 9 t. This distribution happens instantly and automatically through our algorithm. Besides writing the algorithm, we’re not interfering with the distribution. The assigned credit is immediately archived to avoid any sort of double-accounting. We don’t handle this data on a clunky excel-spreadsheet, but it all happens on the fly.


The entire process from collection, sorting, aggregating, and final testing takes time. Your counter on the impact report is usually updated every few weeks and not whenever we register waste at any collection center. Through these processes, we can guarantee that all volumes on your impact report are verified and only relate to absolute dry plastic - precisely what you contracted us for.

If you want to learn more, send us an email at