Sesotec: AI delivers great benefits for plastics recycling
In this VDMA interview on the way to K 2025, Dietmar Dieing, Vice President Sales Plast at Sesotec GmbH, discussed the challenges in plastic recycling and the trend of using artificial intelligence (AI) in plastic recycling.
Dietmar Dieing, Vice President Sales Plast at Sesotec.
Mr. Dieing, how can planning certainty be provided in terms of the circular economy?
Dieing: Planning certainty is extremely important for the success of the circular economy. It can only be achieved if reliable material flows are produced, standardized recyclate quality is obtained, and stable legal framework conditions are in place; these are three prerequisites that currently do not exist. We do not yet have transparent data on products along the entire value chain; a kind of material passport so to speak. We also do not yet have the necessary partnerships between producers, recyclers and machine manufacturers, or political incentives and detailed standards. We will only be able to sort waste through our sorting machines in such a way that a waste product can later correspond to its original product once all of this is in place.
Which of the above requirements is the most difficult to implement?
Dieing: The most difficult are probably the political standards and incentive systems for Europe. These must ensure that everyone works under the same conditions. So far, this has only been fragmentary. For example: there is a requirement from the EU Commission that at least 30% of recycled materials from post-consumer waste must be used in new products as of 2030.
These quotas cannot be met at present, because it is not yet clear exactly in what quality grades the recycled material is available on the market, and in which quantities. It would therefore be important to introduce a regulation stipulating that all products must be manufactured in such a way that they can be recycled.
That means a Design for Recycling?
Dieing: Exactly. Today, many applications still use composite materials that cannot be reasonably converted into recycled materials. Take multi-layer films in the packaging sector for instance: these consist of several different plastics and cannot be recycled in such a way to produce highquality films. Mono-layer films on the other hand are very easy to recycle because they consist of a single polymer. The material flow that is then created begins to be recyclable. The goal must be to achieve the best possible material flow, whereby we can state that everything made from this material is truly 100% recyclable. Designing for recycling purposes must be mandatory. If no specific regulation is set, the whole thing will not work.
When do you expect these specifications to be in place?
Dieing: I am very sure that we still have a long way to go. Nothing is likely to change in the short term. This is due to the economic situation we are facing today, and foreign competition, including in mechanical engineering. The technological possibilities are there, but they are expensive. Many customers are simply not prepared to pay the additional costs.
What are the benefits of artificial intelligence in plastics recycling?
Dieing: I see a huge benefit in this area: in plastics recycling, we have to process large amounts of data extremely quickly in order to optimize processes. AI is exactly the right tool to do this. In the future, there will be more and more specialized applications for AI. Today, we can use cameras and sensors on our conveyor belts in a post-consumer waste stream to identify the
manufacturer of a piece of packaging and the material it is made of. With AI, in the future, we will be able to more or less sort in whatever way we wish. For example, we can retrieve data from the internet about a laptop running on the conveyor belt: what processor it has, what components, or how many rare earths have been used. This will offer considerable advantages in the future, because we will be able to sort for value directly from the first material stream. AI also makes material flows safer. Today, we see many recycling plants catching fire because lithium-ion batteries are passing through the shredder. This will no longer happen with the use of AI.
What is already possible at Sesotec today?
Dieing: We already use AI on our machines to teach them to recognise images, i.e. products. These products are then automatically recognised by every one of our sorters around the world. We also use AI in our sensor technology, which means we can determine with almost 100% certainty whether a product is polyethylene, polypropylene or polystyrene.
How will plastics engineering fare in the future?
Dieing: Mechanical engineering companies will hold their own. Innovation in Europe remains very strong. If we consistently consider recyclability, we will be able to stay ahead of the competition in the future, especially of our competitors in China. However, the Chinese have now learned to build good machines too. We must therefore prepare ourselves for new competitive behavior from Asia. Basically, though, I am very optimistic about our future.
