London-based AI waste analytics company Greyparrot has unveiled Deepnest, an innovative platform billed as the world’s first artificial intelligence system designed to connect brand owners with real-time data on how their packaging performs within waste management systems. This launch addresses a significant challenge in the packaging industry: currently, less than one per cent of waste is manually audited, leaving a critical information gap on actual recyclability.

Deepnest draws on data collected by Greyparrot’s AI camera systems installed in sorting and recycling facilities across more than 20 countries. In 2023 alone, these systems detected over 40 billion waste objects, generating a vast dataset that supports brands like Unilever, Asahi, and Amcor in trialling the platform to optimise packaging design and improve recyclability and recovery rates. This real-world insight is essential as global Extended Producer Responsibility (EPR) regulations increasingly require packaging producers to demonstrate recyclability. In the UK, for example, new EPR rules that took effect in January 2025 mandate that major producers conduct recyclability assessments and report the data by October 2025. Brands are also anticipating similar regulations like virgin plastic taxes across other markets, heightening the platform’s relevance.

Greyparrot’s Chief Operating Officer, Gaspard Duthilleul, explained the motivation behind Deepnest, highlighting a systemic issue: “Too often, packaging that’s theoretically recyclable never makes it through the system as intended. Either it can’t be correctly identified by sorting machines, or it’s made from materials with no viable end market.” Deepnest aims to close this loop by providing brands and waste managers with actionable information, helping design packaging suited for recovery and allowing for a clearer understanding of what performs well in live waste conditions.

The platform aggregates and anonymises data on a national or global level, enabling brands to benchmark packaging recycling performance against competitors and category standards. It also facilitates detailed testing of packaging formats within sub-brands before scaling broadly, identification of design elements that hinder recyclability in key markets, and quantification of the impact of research and innovation efforts aimed at enhancing circularity. For waste management firms, access to comprehensive, global waste data helps improve material recovery, reduce contamination, and enhance sorting efficiency and profitability.

Amy Hooper, Head of Innovation at waste management company Biffa, welcomed the collaboration with Greyparrot and the insights Deepnest could bring. She told industry observers, “Improving valuable material recovery outcomes requires a collective effort across the entire value chain. Collaborating with Greyparrot allows our sector to uncover new layers of insight into material recovery that have the potential to inform more effective packaging and product design, policy, and investment.” Such partnerships underline the importance of bridging the historically fragmented waste data landscape between brand owners and waste operators.

The initiative responds to growing environmental pressures as packaging waste generation in the EU rose by nearly 80 million tonnes from 2009 to 2019, underscoring the scale of the challenge. The upcoming EU Packaging and Packaging Waste Regulation mandates that all packaging be recyclable by 2030, making access to real-world performance data increasingly crucial for compliance and environmental impact.

Traditionally, the packaging industry has relied on laboratory tests and software models to predict recyclability, but the absence of real-world operating data from sorting and recycling facilities has posed a significant limitation. Greyparrot’s AI waste intelligence addresses this gap by delivering granular, live data on packaging performance within complex waste streams. Mark Roberts, Circular Economy Director at Amcor, noted: “The packaging industry relies on lab-scale testing and software models to predict recyclability of packaging solutions, but actual real-life data is missing. Deepnest is unlocking real-world recyclability data that the packaging data chain has been missing.”

Asahi Beverages, one of the early adopters, uses Greyparrot’s AI analyzers to enhance operational data quality and inform sustainable packaging efforts. Sandra Gibbs, Asahi’s Chief Supply Chain Officer, highlighted the company’s progress towards sustainable packaging, including switching to 100 per cent recycled plastic bottles for some major brands and operating Australia’s largest PET recycling facility. She stated, “Deepnest can transform that data into insights to guide smarter packaging design from the outset. We’re exploring how this technology can help embed a data-driven approach across the entire packaging lifecycle, moving us closer to 100 per cent circular packaging.”

Greyparrot’s technology also aligns with broader automation trends in waste management. With increasing labour shortages and mounting pressure on Material Recovery Facilities (MRFs) to improve sorting efficiency, AI-powered systems like Greyparrot’s analyzers are being deployed worldwide. These systems provide real-time insights into the composition and quality of incoming waste, enhancing sorting accuracy and operational efficiency while supporting circular economy goals.

Unilever’s Global R&D Head of Deodorants, Dr Liz Smith, emphasised the potential of AI-enabled waste intelligence tools to provide new visibility into how packaging is sorted and processed in real systems. She commented, “Our goal is to reduce our virgin plastic use and make our plastic packaging reusable, recyclable, or compostable — and insights like these could critically help to inform future packaging design, enable recyclability in practice and at scale, and increase the supply of high-quality recycled materials.”

Greyparrot’s Deepnest platform represents a significant step towards closing the information gap in packaging recycling. By delivering real-world performance data, the system helps brands design packaging that not only meets regulatory demands but also meaningfully advances circularity and resource recovery across the global waste value chain.

📌 Reference Map:

  • Paragraph 1 – [1], [4]
  • Paragraph 2 – [1], [2]
  • Paragraph 3 – [1]
  • Paragraph 4 – [1]
  • Paragraph 5 – [1]
  • Paragraph 6 – [1], [5]
  • Paragraph 7 – [1]
  • Paragraph 8 – [1]
  • Paragraph 9 – [1]
  • Paragraph 10 – [1]

Source: Noah Wire Services