The economics of recycling have always been a problem of purity. It’s one thing to sort a clean stream of PET bottles. It’s another to pick a broken piece of treated lumber from a pile of concrete, dirt, and drywall. That’s where the money is, and the human labor is most expensive, hazardous, and hard to find. Eric Camirand, founder of Waste Robotics, has spent eight years teaching robotic arms to see the difference.
His company builds AI-powered sorting robots for recycling and waste facilities, a category that is less about gleaming future factories and more about retrofitting noisy, dusty buildings where the work is fundamentally unpleasant. The bet is that computer vision and deep learning have finally gotten good enough, and cheap enough, to automate the sorting of heterogeneous waste streams like construction and demolition (C&D) debris and organic waste. The robots use off-the-shelf arms but pair them with proprietary vision systems, including hyperspectral cameras, to identify and extract materials as they move down a conveyor belt [Waste Robotics, retrieved 2024].
A wedge in construction debris
Waste Robotics didn't start with the easy stuff. While many robotic sorters target municipal recycling facilities with relatively clean streams of cans and bottles, the company's early focus was on C&D waste and organics. This is a classic wedge: the material is more variable and contaminated, making it harder for traditional optical sorters, but the manual alternative is so costly and unattractive that customers have a stronger incentive to automate. The company claims its systems can achieve up to 5% higher purity rates by using hyperspectral cameras, which analyze light wavelengths beyond the visible spectrum, compared to standard AI vision [Waste Robotics, retrieved 2024]. This lets a robot distinguish, for instance, between different types of plastics or between natural wood and pressure-treated lumber coated in chemicals.
Their business model blends hardware and software. They sell the robotic cells, but also offer a "Sorting as a Service" program, which likely helps mitigate the high upfront capital cost for facility operators [Waste Robotics, Mar 2021]. The primary customers are municipal and private material recovery facilities (MRFs) and processors who want to increase throughput, reduce labor costs, and improve the quality of their output commodity bales.
Funding a transatlantic push
In September 2023, Waste Robotics secured a CAD 10 million (approximately USD 7.3 million) Series A round led by French impact investor Mirova and Canadian fund Fondaction [SPEEDA Edge, Sep 2023]. The capital was earmarked for expansion into France, the UK, and deeper into North America, signaling a move from proving the technology to scaling deployments. The round brought the company's total disclosed funding to roughly USD 7.4 million [PitchBook, 2025].
2019 Seed | Undisclosed
2023 Series A | 7.4 | M USD
This institutional backing from sustainability-focused funds like Mirova and Fondaction is a vote of confidence in the unit economics of robotic sorting. The investment case isn't just about selling robots; it's about enabling a more efficient and profitable circular economy for materials that are currently downcycled or landfilled because they're too expensive to sort.
The team and the traction
Founder and CEO Eric Camirand has led the company since its 2016 founding in Trois-Rivières, Quebec, building a team that now numbers between 11 and 50 people according to LinkedIn [LinkedIn, retrieved 2024]. The leadership includes VP of Products Ziad Akl-Chedid and founding partner Michel Laforest [Specim, Spectral Imaging Ltd. on LinkedIn, 2026]. While the company keeps specific customer names close, trade press points to deployments like the installation of two heavy sorting robots with "Gripper AI" technology at Millennium Recycling [Recycling Inside, 2026]. They've also partnered with camera manufacturer Specim to integrate hyperspectral imaging, a technical collaboration that serves as a signal of their focus on advanced material discrimination [Specim, Spectral Imaging Ltd. on LinkedIn, 2026].
Where the wheels could come off
The path for Waste Robotics is not without its potholes. The market, while growing, is competitive and capital-intensive.
- The scaling challenge. Selling heavy industrial hardware into waste facilities is a slow, relationship-driven sales cycle. The stated expansion into Europe will require building new sales and service networks from a Canadian base, a costly and complex undertaking.
- Technical limits. Even with hyperspectral imaging, certain material distinctions remain incredibly difficult for any vision system,think black plastics or heavily soiled items. The company's 5% purity claim is a relative improvement, but the absolute accuracy in real-world, dirty conditions is the metric that matters for customer payback.
- Competitive pressure. They are not alone. Competitors like Ireland's Danu Robotics are also targeting the waste sorting space with AI and robotics [Tracxn, 2026]. The longer-term threat may come from the large waste management conglomerates developing in-house automation solutions, though their pace of innovation is typically slower.
The company's most plausible answer to these risks is its focused wedge. By specializing in the dirtiest, most complex waste streams first, they build a defensible moat of data and expertise that a generalist robot maker or a slow-moving incumbent would find hard to replicate quickly.
The math on a sorting line
Let's do a back-of-the-envelope check. A single manual sorter in a North American facility might cost a company $50,000 to $70,000 per year in wages and benefits. A robotic sorter from Waste Robotics likely costs several times that upfront but can operate for 20 hours a day without breaks. If one robot can replace two full-time positions, the payback period could sit in the three-to-five-year range, before accounting for the value of increased purity and throughput. That's the fundamental equation the company is selling: capex for opex, with a side of consistency.
For Waste Robotics to graduate from a promising hardware vendor to a category-defining company, it must beat the incumbent that isn't another startup, but the status quo: the temporary labor agency. It must prove that its robots are not just a novel piece of equipment, but a more reliable and ultimately cheaper employee that shows up every day and doesn't mind the dust. If they can do that in the gritty world of C&D yards, the cleaner streams of tomorrow's recycling plants will look like a straightforward next step.
Sources
- [Waste Robotics, retrieved 2024] Company website and product claims | https://wasterobotic.com/
- [SPEEDA Edge, Sep 2023] Funding round announcement | https://sp-edge.com/companies/669933
- [PitchBook, 2025] Funding history and investor details | https://pitchbook.com/profiles/company/184397-50
- [LinkedIn, retrieved 2024] Company headcount estimate | https://ca.linkedin.com/company/wasterobotics
- [Waste Robotics, Mar 2021] Sorting as a Service mention | https://wasterobotic.com/2021/03/17/fondaction-invests-in-waste-robotics/
- [Specim, Spectral Imaging Ltd. on LinkedIn, 2026] Partnership announcement | https://www.linkedin.com/company/specim-spectral-imaging-ltd/posts/?feedView=all
- [Recycling Inside, 2026] Millennium Recycling deployment | https://recyclinginside.com/robotics-automation/waste-robotics-deploys-two-heavy-sorting-robots-at-millennium-recycling/
- [Tracxn, 2026] Competitor information | https://tracxn.com/d/companies/waste-robotic/___1_TzNZbJPXtKK4IQcaO99rW7fxb7EZHS2lMYoU4998