The economics of recycling are, at their core, a sorting problem. A materials recovery facility (MRF) makes money by selling clean bales of plastic, paper, or metal, but the conveyor belts delivering that material are a chaotic, high-speed mess. For Arthur Goujon and Luis Espinosa, the co-founders of London's Sorted Technologies, the answer wasn't to rip out the belts and install expensive robotic arms. It was to give the human pickers standing beside them a better set of eyes and a laser pointer.
Sorted builds AI-powered vision systems that use computer vision and spectroscopy to identify specific materials on a sorting line. Then, it projects a colored laser dot directly onto the target item,a green dot for a PET bottle, a red dot for an HDPE container,telling the picker exactly what to grab. The company's recent €1.9 million (about $2.05 million) seed round, led by Pi Labs with participation from Archipelago Ventures and others, is a bet that this human-in-the-loop, retrofittable approach is the wedge into the capital-constrained, slow-moving world of waste management [EU-Startups, April 2024].
The retrofit wedge
The sorting robotics market has no shortage of well-funded contenders, from AMP Robotics in the US to Greyparrot and Recycleye in Europe. Their value proposition is clear: robots don't get tired, they sort with inhuman consistency, and they work in environments humans shouldn't. The problem is they often require a significant capital outlay and physical integration that can disrupt existing operations. Sorted's pitch is different. Its 'Sorted Light' system is a sensor array and projector that mounts above an existing conveyor line. It doesn't replace the picker; it augments them.
- Lower barrier to entry. By avoiding the mechanical complexity and cost of a robotic arm, Sorted aims for a lower CAPEX solution that facility managers can approve without a multi-year ROI calculation.
- Operational continuity. The system is designed to be installed and calibrated with minimal downtime, a critical factor for facilities running multiple shifts.
- Data first, automation later. Sorted also offers a 'Waste Audit Studio' that uses image analysis to generate waste composition reports, and a compact robotic sorter called 'SortBot' for specific tasks. The portfolio allows customers to start with analytics and guided picking, then scale into partial automation [sortedtech.io].
Founders from both sides of the belt
The co-founding team brings together the two domains that must meet for Sorted to work: industrial operations and software. Luis Espinosa is described in coverage as a 'waste pioneer' with a background in the recycling sector, providing the ground-level understanding of MRF workflows and pain points [TechFundingNews, November 2023]. Arthur Goujon, the CTO, was previously Head of Technology at The Very Group, a major UK retailer, where he led digital and logistics technology teams [EU-Startups, April 2024]. This combination,one founder who knows the garbage, the other who knows the code,is a classic pattern for industrial tech startups trying to bridge a physical-digital gap.
Early traction and the metrics question
Publicly, Sorted names a handful of early customers, including waste management giants SUEZ and NWH Group, as well as contractor Bywaters [sortedtech.io]. The company's website cites impressive efficiency gains from these deployments: a 17% efficiency increase for NWH Group in a month, and a 97% reduction in waste audit time for Bywaters. These figures, while compelling, are common in early-stage case studies and serve as the necessary proof points to secure seed funding and attract the next wave of customers. The real test will be whether they can be replicated and validated at scale across dozens of facilities with varying input streams.
A crowded field of robots and cameras
The risk for Sorted is that its wedge,being cheaper and less disruptive,might also be easier to bypass. The competitive landscape is dense with companies pushing the automation envelope.
| Competitor | Primary Focus | Geography | Notable Differentiation |
|---|---|---|---|
| AMP Robotics | Full robotic sorting | Global (US-based) | Extensive deployment history, proprietary AI platform |
| Greyparrot | AI waste analytics | Europe | Strong focus on software and data for facility optimization |
| Recycleye | Robotic picking | Europe | Computer vision for high-value material recovery |
| Bulk Handling Systems | Turn-key sorting systems | Global | Full-system supplier for large, new facilities |
Sorted's answer is that full robotic replacement is a decade-long project for most existing MRFs, and in the meantime, there is immense value in making the current human-powered system radically better. Their focus on retrofit and augmentation carves out a specific niche: the vast middle layer of facilities that need to improve yields now but cannot or will not commit to a full robotic overhaul.
The path to proving unit economics
The next twelve months for Sorted will be about moving from pilot projects to repeatable deployments. The seed funding is earmarked for scaling deployments and building out the team [EU-Startups, April 2024]. The key metric to watch won't be the percentage efficiency gain in a single facility, but the number of facilities running Sorted systems on multiple shifts, day in and day out. Each installation is a case study in unit economics: how much did recovered material value increase, and how did that stack up against the system's cost?
A back-of-the-envelope calculation illustrates the bet. If a typical picking station handles 2 tons of material per hour and a Sorted system increases recovered value by even 10%, that could mean an extra $20-40 of revenue per hour, depending on material mix. Over three shifts, that adds up. The system isn't competing with a $200,000 robotic cell on pure performance; it's competing on a faster payback period and less operational friction.
Sorted's ultimate incumbent to beat isn't another startup. It's the status quo of a worker staring at a fast-moving belt, making split-second decisions based on glance and guesswork. If their colored lasers can turn that guesswork into a sure thing, the economics of the entire belt start to change.
Sources
- [EU-Startups, April 2024] London-based recycling startup Sorted raises over €1.9 million, aiming to solve material sorting gap | https://www.eu-startups.com/2024/04/london-based-recycling-startup-sorted-raises-over-e1-9-million-aiming-to-solve-material-sorting-gap/
- [TechFundingNews, November 2023] From a pub chat to a million-pound startup: How Sorted are revolutionising recycling | https://techfundingnews.com/from-a-pub-chat-to-a-million-pound-startup-how-sorted-are-revolutionising-recycling/
- [sortedtech.io] Discover Sorted's AI-powered recycling technology | https://www.sortedtech.io/
- [SeedTable, 2024] Sorted Company Information - Funding, Investors, and More | https://www.seedtable.com/startups/Sorted-9A5AJD3