IPercept raises $16.5M Series A to give CNC machines a health check

Stockholm’s IPercept has raised a $16.5M Series A co-led by Isogon Ventures and 2150, with Luminar Ventures, RunwayFBU, J12 Ventures and AI.Fund participating, to take its CNC machine health monitoring into the US.

Europe keeps hearing that it missed the software platforms and is about to miss the AI ones too. And yet the machines that make every physical thing around you – the CNC machine tools cutting metal for aerospace, automotive and energy – still sit on European factory floors, and this morning a Swedish company raised proper money to watch them more closely. Below, I lay out the round, the technology behind it, and what it tells us about where industrial capital is flowing.

Two co-leads and a ticket across the Atlantic

IPercept announced its $16.5M Series A on 29 September 2026, co-led by Isogon Ventures and 2150, the London-based sustainability investor. Luminar Ventures, RunwayFBU, J12 Ventures and AI.Fund – all backers from the company’s earlier rounds – joined again, which is its own kind of signal.

The company says the capital goes to a US market launch with its first local hires there, continued product development, and deeper partnerships with the industrial service companies that deploy its hardware. The round follows a €5M seed in May 2025 led by Luminar Ventures, and takes the Stockholm company from Nordic proof to transatlantic ambition in under eighteen months.

An MRI for the machines that make machines

IPercept spun out of KTH Royal Institute of Technology, was founded in 2019 and has sold commercially since 2022, with a team the company puts at just over 20 people. Its device clamps onto the moving section of a CNC machine and measures motion at micrometre precision, catching wear and degradation in bearings, ballscrews, guides and spindles. Crucially, it never touches the machine’s controller or the plant’s IT network – which is exactly why security-conscious manufacturers let it through the door.

“CNC machine tools make every other machine on Earth, and until now almost nobody could see inside them while they run,” founder and CEO Karoly Szipka said in the announcement, comparing the readout to a full-body MRI scan for large industrial machines. The customer list backs the claim: Airbus, Bosch, Scania, Hitachi Energy and Volvo use the system, alongside industrial service partners such as Konecranes and Quant.

A market the analysts cannot agree on

How big is the prize? Honest answer: the analysts disagree by an order of magnitude. Mordor Intelligence sizes predictive maintenance at $18.9bn in 2026, growing at a 34.14% CAGR to $82.17bn by 2031, while Research and Markets, in a September 2026 report, puts the same 2026 market at $9.71bn with an 11.5% CAGR to 2031. The spread is itself informative: nobody doubts the direction, only how fast factories will pay for foresight. What is not in dispute is the installed base – the company puts the value of CNC machines in use worldwide at around $1 trillion.

For scale against our own fundraising data: London’s CloudNC raised $20M this month, led by Nimble Ventures, to automate the programming side of the same CNC world, and earlier industrial IoT rounds we tracked – Forgis’s €3.8M for manufacturing automation and CERPRO’s €2M for industrial workflows – were seed-sized bets on the same thesis. A mid-teens Series A puts IPercept exactly where a category leader in European industrial tech should be after four years of commercial traction.

What a Swedish sensor round signals

Three patterns are worth watching here. First, the university pipeline works: KTH metrology research became a venture-scale company the same way ETH and TU Munich spinouts have been doing across deep tech – slowly, then credibly. Second, the buyer profile is the story. When Airbus, Bosch and Volvo adopt a 20-person startup’s hardware, the constraint is no longer product-market fit; it is deployment capacity, which is exactly what the service-partner strategy addresses.

So why does a sustainability fund co-lead a machine-tool round? Because uptime is a climate lever – a machine that fails less is a factory that scraps less and consumes less – and 2150 betting on Swedish sensors says industrial efficiency has quietly become a climate thesis. The good news is that Europe is structurally long on exactly this: dense manufacturing, strong technical universities, and founders who have stood next to the machines they now instrument. The reality remains that the expansion market is the US, and the race is whether Nordic industrial AI can hire in America faster than American rivals can copy Swedish hardware.

Watch the quiet ones. While the loud end of AI fights over chatbots, companies like IPercept are instrumenting the machines everything else depends on – and that is the layer Europe can actually own. Let’s keep backing it.