Jaipur Robotics raises €4.3M to give waste-to-energy plants an AI operating system

Jaipur Robotics (Manno, Switzerland) has closed a €4.3 million Seed round co-led by EquityPitcher Ventures and High-Tech Gründerfonds (HTGF). The company, founded in 2024 by Ermes Zamboni and Nikhil Prakash, builds computer vision and AI systems for waste-to-energy plants, cement facilities, and biomass sites — markets where the cost of an unplanned shutdown can run to hundreds of thousands of euros, and where the operational workforce shortage is structural.

The round follows a €725,000 pre-Seed led by TiVentures (March 2025) and an additional €161,000 from Venture Kick. The 40-person team operates out of Technopole Ticino in Manno, near Lugano, with a second R&D centre in Asia.

The dataset as the product

Waste-to-energy plants share a specific and expensive problem: non-recyclable waste arriving from municipal collections is heterogeneous by nature, and hazardous objects — gas cylinders, concrete blocks, oversized metal — hidden within incoming loads cause combustion chamber failures when they are not caught before entering the furnace. An unplanned shutdown costs significantly more than prevention. Historically, crane operators have managed this manually: scanning camera feeds, directing the crane to sort hazardous material from the bunker, and making combustion judgements based on experience.

Jaipur Robotics applies computer vision to the process. Its system detects oversized and dangerous objects in bunkers and truck unloading bays in real time and maps the lower heating value (LHV) — the calorific content — of waste across different bunker zones, presenting crane operators with a heat map that shows where high- and low-energy waste is concentrated. Mixed correctly before combustion, the load optimises energy output and reduces emissions. The company also tracks crane position and operational status in real time for predictive scheduling.

The results the company reports are specific: more than 80% fewer unplanned shutdowns from hazard detection; more than €1 million in added annual value per plant from improved waste mixing; crane operations that are measurably more productive. The underlying asset that makes those numbers possible is not a better model — it is the company's 50 million labelled images of waste, built from operating plants over two years. That corpus is what a competitor would have to replicate before they could begin to match the accuracy Jaipur reports.

Expansion and the industrial sales question

With the Seed capital, Jaipur Robotics is targeting new geographies and additional industrial plant types beyond waste-to-energy. The mention of India in early coverage of the round — the company's Asia R&D presence and a founder's background — suggests that Indian waste infrastructure is on the expansion roadmap.

Anna Stetter of HTGF — a German fund's decision to back a Swiss company expanding into waste-to-energy markets is itself a signal about where they see the demand — named the combination of data, team, and verifiable results as the investment thesis: "What convinced us to co-lead this round is the combination of a dataset, a founding team with rare depth across mechanical engineering, deep learning, and industrial deployment, and outstanding, measurable results at customer sites."

Ermes Zamboni, CEO: "We envision a future where data from waste fully drives safety and automation. This round enables us to revolutionise entire plant operations."

The product numbers are real. The question is whether WtE plant operators sign and deploy fast enough to justify the capital. Industrial procurement cycles are long; plant managers are conservative; the switching cost of adding a new system to an operating facility is non-trivial. Jaipur's task is not to prove the technology — the results at existing sites do that — but to compress the sales cycle in markets where it does not yet have reference customers.