IPercept, the company whose patented technology enablesCNC (Computer Numerical Control) machine tools to report their own mechanicalcondition, has raised a $16.5m Series A round, co-led by Isogon Ventures and 2150. Existing investors Luminar Ventures, RunwayFBU, J12 Ventures, and AI.Fund all participated in the round.
The Series A funding provides IPercept with the capital to launch and scale operations in the US and make its first dedicated local hires. IPercept will also extend its data platform with a series of add-on intelligence applications - the software layer the company believes will make physical AI practical for discrete parts manufacturing.
IPercept has built the data layer that can unlock physical AI for the world’s largest existing robotics segment: its plug-and-play technology unlocks uptime, productivity, and quality, while engaging teams via an all-in-one CNC machine intelligence suite.
Already live across facilities for over 30 manufacturers including Airbus, Bosch,Hitachi Energy, Scania, and Volvo, IPercept unlocks peak machine performance and allows better decisions on high-quality data analysis, producing a value loop of less unplanned downtime and improved overall equipment effectiveness. IPercept already has traction in the industrial service company sector, with customers including Konecranes, Quant,DynaMate, and MTT deploying the technology as the basis of their own monitoring services.
Operating as an AI assistant for production and machine maintenance, IPercept's AI-driven analytics solutions for CNC machines give a level of detail not previously available to manufacturers - clearing an industrial blind spot that had existed since the 1940s, now solved by the KTH Royal Institute of Technology spinout.
IPercept’s technology reads the machine’s own motion and a single patented IPercept device mounted on the moving part of a CNC machine (any brand, model, or age) converts micrometre-scale motion into component-level condition data, with no connection to the machine controller or the customer’s IT network. Manufacturers see which components are wearing, how quickly, and exactly what to do before productions tops: until now, operators did not have this visibility.
CNC machines are inputs into huge industries (e.g. automotive, aerospace, electronics, medical devices, machinery, energy, defence) and critical to the global economy. CNC monitoring stood still for 25 years as no scalable data access for these machines existed. For companies not running IPercept technology, it still takes two days of manual work to fully understand why a single machine failed, while average machine efficiency has remained stuck at just 50–60% for decades.
Industry needsto quickly shift the world’s manufacturing centre of gravity back to the West,shaped by ever changing geopolitics and commercial objectives. At the centre ofthat problem sit CNC machine tools - the precision systems used to manufacturethe components that underpin modern life, from aircraft engines and medicalimplants to cars, electronics and energy infrastructure - with an installedbase worth roughly $1 trillion worldwide.
They are alsothe least observed asset in the factory: the machine controller commands motionbut does not measure mechanical condition, and vibration sensors were designedfor pumps and motors, not machines with dozens of interacting components.
IPerceptfounder and CEO Karoly Szipka:“CNC machine tools make every other machine on Earth, and until now almostnobody could see inside them while they run. The question we set out to answerwas how to make the most important machine in the physical world legible tosoftware and AI. That took a new measurement principle, five years of research,and four years of real-world data. We are ready to bring it to every factorythat needs it.”
.jpg)
