The water in an aeration basin is a living, breathing ecosystem, a microbial city working to break down waste. For the operators tasked with keeping it healthy, the primary diagnostic tool has remained unchanged for decades: a sample bottle, a courier van, and a lab report that arrives days later, describing a problem that may have already bloomed. Lir Labs, a London-based company founded last year by four graduates of a joint Royal College of Art and Imperial College design engineering program, is betting that timeline is a relic. They are building deployable, AI-powered microscopes to give wastewater plants and water utilities a real-time read on everything from dangerous asbestos fibers to the shifting balance of biological life.
Their wedge is speed, translating a process measured in days into one measured in minutes. The promise, as highlighted in a university post, is reducing water testing time "from days to minutes" compared to traditional lab methods [Royal College of Art Facebook, 2025]. For an operator, the difference is between reactive triage and proactive management. A secondary source claims the company built "the first real-time OpticalAI sensor for asbestos in water," directly linking contamination levels to pipe degradation [Prospeo, Unknown]. It is a classic hardware-plus-software play: the microscope gathers continuous optical data on particle morphology and species distribution, and the proprietary AI models parse it into actionable alerts on an industrial dashboard.
The design engineering wedge
The founding team's background is their first differentiator. All four co-founders,Diego Munoz Iglesias, Emile Meunier, Benedikt Huber, and Doruk Tan Ozturk,are graduates of the Innovative Design Engineering MA, a program known for blending technical rigor with user-centric design [Royal College of Art Facebook, 2025]. This isn't a team of pure microbiologists or optical physicists; it's a group trained to build tangible systems that solve messy, real-world problems. The product they describe is less a lab instrument and more an industrial sensor, meant to be deployed in the humid, vibrating environment of a treatment plant, not a sterile bench. Munoz Iglesias holds majority control of the company [UK Companies House, retrieved 2024]. Their early focus on a critical, regulated contaminant like asbestos suggests a strategy of targeting a high-stakes, unambiguous pain point to gain an initial foothold.
Where the unit economics get real
The financial logic for a water utility is straightforward, if you run the numbers. A traditional lab test for a suite of contaminants can cost hundreds of dollars per sample and take three to five business days. For a plant running multiple tests per week, the annual bill is significant, but the larger cost is operational risk. A process upset that goes undetected for days can lead to regulatory fines, equipment damage, or costly remediation. Lir Labs would presumably charge a recurring fee for its hardware-as-a-service and software dashboard. If that subscription cost is less than a utility's current annual spend on outsourced lab testing,and provides continuous data instead of periodic snapshots,the value proposition clicks. The real savings, however, are in avoided catastrophes. Spotting a spike in filamentous bacteria early can prevent a settling tank from failing. Detecting asbestos fibers could trigger pipe maintenance before a costly rupture. The company's bet is that operators will pay not just for data, but for the insurance policy of perpetual vigilance.
The incumbent is the courier van
The most direct competitor isn't another startup; it's the entrenched workflow of sample bottles and accredited external laboratories. This ecosystem is built on standards, regulations, and trust built over decades. For Lir Labs to displace it, their AI-driven analysis must achieve a level of accuracy and regulatory acceptance that matches or exceeds the human-in-the-loop lab. Their early claims are bold, but the path to widespread adoption requires rigorous validation and potentially years of side-by-side correlation studies to prove their optical readings are as reliable as chemical assays.
Other risks on the horizon include:
- Technical complexity. Monitoring a dynamic biological ecosystem is harder than detecting a single contaminant. The AI models must be robust against changing water compositions, weather events, and seasonal shifts.
- Sales motion. Selling capital hardware into slow-moving, budget-conscious municipal utilities is a famously long cycle. The company's pre-seed backing of approximately $266,700 from Antler is a start, but it is a fraction of what is typically required to fund a sales engineering team and pilot deployments [Preqin, 2025-2026].
- Defensibility. The core innovation sits at the intersection of optical hardware and machine learning. Both are fields with rapidly falling tooling costs. Their lead may depend on the proprietary dataset they build from early deployments and the specific industrial design of their deployable units.
A snapshot of the founding team shows their shared academic origin and early roles:
| Founder | Role / Background Note |
|---|---|
| Diego Munoz Iglesias | Controlling founder (75%+ voting rights) [UK Companies House, retrieved 2024]. |
| Emile Meunier | Co-founder, Head of Product; studied at Imperial College London [LinkedIn, retrieved 2026]. |
| Benedikt Huber | Co-founder, IDE program graduate [Royal College of Art, 2025]. |
| Doruk Tan Ozturk | Co-founder, Materials Engineer and Interdisciplinary Designer [Intch, retrieved 2026]. |
The next twelve months
The immediate task for Lir Labs is to move from prototype to proven pilot. The pre-seed capital is for building the first commercial-grade units and getting them into the hands of a reference customer,likely a forward-thinking municipal plant or an industrial water user. Traction will be measured not in website clicks, but in the number of basins where their sensors are installed and the months of continuous, validated data they collect. The dashboard they preview online, with its graphs for "Ecosystem Quality" and "System Resilience," must evolve from a demo into a daily decision-making tool for plant managers [Lir Labs HMI Dashboard, retrieved 2026].
Back-of-the-envelope, if a single microscope unit can monitor a key basin and replace 52 lab tests per year at an average cost of $500 each, it saves $26,000 annually in direct testing fees. If the subscription is priced at, say, $15,000 per year, the utility nets an $11,000 saving while gaining real-time insight. The bigger win is harder to quantify but more valuable: turning a blind, delayed process into a visible, manageable one. For Lir Labs to succeed, they must prove their system is not just faster, but as trustworthy as the lab report that arrives in a courier's van. That is the incumbent they have to beat.
Sources
- [UK Companies House, retrieved 2024] LIR LABS LTD overview and persons with significant control | https://find-and-update.company-information.service.gov.uk/company/16554279
- [Preqin, 2025-2026] Preqin asset profile for Lir Labs | https://www.preqin.com/data/profile/asset/lir-labs/799601
- [Royal College of Art Facebook, 2025] RCA post on Lir Labs founders and technology | https://www.facebook.com/RoyalCollegeOfArt.RCA/posts/-congratulations-to-lir-labs-who-will-represent-europe-this-friday-in-the-global/1196734969164132/
- [Prospeo, Unknown] Prospeo company overview citing asbestos sensor | https://prospeo.io/c/lir-labs-revenue
- [LinkedIn, retrieved 2026] Emile Meunier LinkedIn profile | https://ca.linkedin.com/in/emile-meunier-096592259
- [Royal College of Art, 2025] Benedikt Huber student profile | https://2025.rca.ac.uk/school-of-design/innovation-design-engineering-ma-msc/profile/benedikt-huber/
- [Intch, retrieved 2026] Doruk Tan Ozturk profile | https://intch.org/Toto
- [Lir Labs HMI Dashboard, retrieved 2026] Wastewater Treatment Plant Dashboard preview | https://lirlabs.live/