Artificial Intelligence: Smart Solutions for Plastic Recycling

At the Future Packaging Day 2025 in Oberwaltersdorf, Volker Muche, co-founder of Pacoon Sustainability Concepts, posed a central question: How can artificial intelligence (AI) advance the circular economy in the packaging sector? His keynote “Smart Solutions: How AI is Transforming Plastic Recycling” highlighted the opportunities and challenges of this technology.

Muche began with an appeal: The era of the throwaway economy is over. Packaging must not only be functional but also sustainable. Yet reality looks different. Muche provided examples, such as the fact that only about 9% of plastic waste is currently recycled. At the same time, the PPWR regulation requires that packaging must contain a minimum share of recycled material in the future. The industry therefore faces a dual challenge: ensuring quality and closing loops.

Muche emphasized: AI is not an abstract buzzword but a tool that can transform the entire value chain, from design to sorting to sourcing recycled materials.

A practical example is the research project KIOptipack, which Muche presented. The goal: to optimize the development and production of packaging with a high recycling content through the use of AI. All process steps: design, material selection, manufacturing, sorting, and reprocessing, are networked in a shared data space. AI not only supports material selection but also simulates recycling scenarios and evaluates ecological and economic impacts. This creates a holistic approach to sustainable packaging.

Furthermore, artificial intelligence can revolutionize sorting plants. Material separation is still error-prone, black bottles or complex composites often end up in the wrong stream. AI can make the difference here: through deep learning and millions of image data, it learns to recognize packaging precisely and evaluate it in real time. Intelligent nozzles or robots then sort with maximum accuracy. The result: higher purity, less waste, more valuable material.

In addition to sorting, packaging development plays a central role. AI can act as a “co-pilot” here: it creates digital twins of packaging, simulates transport, usage, and recycling processes, and detects design flaws early. Developers receive immediate feedback on the ecological impact of their decisions. This saves time, reduces costs, and accelerates innovation cycles. However, Muche also emphasizes that reality is often more complex than theory:

An often underestimated problem: the smell of recyclates. Today, evaluation is mostly done by humans, a bottleneck that will worsen as recycling rates increase. AI-based systems can help by objectively measuring and predicting odors.

Opportunities and Limits
The advantages are clear:

• Smarter design through AI-supported development
• More stable processes thanks to real-time quality control
• Transparent sourcing of recyclates via digital platforms

But Muche also addressed the stumbling blocks: data quality, data protection, high implementation costs, and ethical questions. And last but not least: AI is only as good as the data that feeds it. How do we ensure that it not only reproduces existing patterns but enables true innovation?

Muche made it clear: AI is not an end in itself. It is a tool that helps us master the complexity of the circular economy: faster, more efficiently, smarter. But the path remains challenging. It requires investments, cooperation, and the courage to take new paths.