IPercept Raises $16.5M Series A to Bring Physical AI to CNC Machines

Stockholm-based industrial AI startup IPercept has raised $16.5 million in Series A funding to expand its physical AI technology for CNC machines and industrial manufacturing.

The round was co-led by Isogon Ventures and 2150, with continued participation from Luminar Ventures, RunwayFBU, J12 Ventures and AI.Fund.

Founded in 2019 by Karoly Szipka and Andreas Archenti, IPercept develops a machine diagnostics and predictive maintenance system designed to monitor CNC machines without accessing their internal controllers or connecting to a customer's IT network.

The company says its technology can provide detailed insight into the condition and movement of machine components, helping manufacturers identify potential problems before they result in costly unplanned downtime.

IPercept currently works with more than 30 enterprise customers, including Airbus, Bosch, Hitachi Energy, Scania and Volvo. The company has grown to more than 20 employees and plans to use the latest financing to expand its partnerships with industrial service companies and enter the US market.

Bringing physical AI to CNC machines

CNC, or computer numerical control, machines are central to modern manufacturing. They are used to precisely cut, shape and manufacture components for industries ranging from automotive and aerospace to energy and industrial equipment.

Despite their importance, CNC machines have historically operated with relatively limited monitoring.

IPercept says average CNC machine efficiency has remained around 50-60% for decades.

That creates a significant opportunity for technologies that can help manufacturers understand what is happening inside their machine fleets and identify issues before they become failures.

Rather than focusing on humanoid robots or warehouse automation, IPercept is targeting the machines already operating inside factories.

The company's approach is based on specialized physical AI and machine diagnostics.

Karoly Szipka, IPercept's co-founder and CEO, has a PhD in CNC machine diagnostics from the KTH Royal Institute of Technology and has worked in the field for more than a decade.

He co-founded IPercept with Andreas Archenti, a chair professor at KTH specializing in industrial dependability.

The company was established in 2019 and brought its technology to market in 2022.

A domain-specific approach

Szipka says IPercept's technology was developed after recognizing that monitoring CNC machines was not simply a conventional data science problem.

Swedish manufacturers had initially attempted to address CNC monitoring by bringing in data scientists.

According to Szipka, the company eventually determined that the challenge required deep domain knowledge about industrial machinery before applying AI and data analysis.

That insight became central to IPercept's product development.

Instead of relying primarily on large historical datasets, the company's technology focuses on directly measuring the physical behavior of the machine.

The company says its solution does not require historical machine data.

That allows it to be deployed even where manufacturers do not have extensive historical records of machine performance.

An MRI-like approach to machine diagnostics

IPercept's hardware can be attached to the moving section of a CNC machine.

The device is designed to work across different brands, models and generations of machines.

According to the company, it can measure movement at the micrometre level without connecting to the machine controller or the customer's IT network.

Szipka compares the approach to an MRI scan.

The idea is to create a detailed understanding of the machine's physical condition and identify abnormalities before they lead to breakdowns.

The system can therefore operate independently of the machine's existing control architecture.

That distinction is important for factories with large fleets of machines from different manufacturers.

IPercept says its technology is compatible with around 300 CNC machine manufacturers and can be installed on an individual machine within a few hours.

The company says the world's installed CNC machine base is worth approximately $1 trillion.

Predictive maintenance without controller access

Traditional industrial monitoring approaches can involve connecting directly to machine controllers or installing sensors designed to monitor specific components.

IPercept takes a different approach.

Its system measures machine motion directly, allowing it to assess the behavior of the machine without requiring access to the controller.

That can simplify deployment, particularly for manufacturers operating older equipment or machines from different vendors.

The company positions the technology as a predictive maintenance tool that can identify potential failures before they cause production interruptions.

Unplanned downtime can have significant consequences for manufacturers because a single machine failure can interrupt production schedules, delay deliveries and require emergency maintenance.

By providing earlier insight into machine condition, predictive maintenance systems can potentially allow manufacturers to schedule interventions rather than responding to unexpected breakdowns.

The broader predictive maintenance market is projected to reach $97.37 billion by 2034, according to the market research cited in the source.

More than 30 enterprise customers

IPercept has built a customer base across major industrial companies.

Its customers include Airbus, Bosch, Hitachi Energy, Scania and Volvo.

The company also works with industrial service companies such as Konecranes, Quant, DynaMate and MTT.

These service providers can resell IPercept's monitoring technology as part of their maintenance agreements.

That channel strategy gives the company another route into industrial customers beyond direct sales.

IPercept says it now has more than 30 well-known enterprise customers and that the business has grown more than threefold during the previous six to eight months.

The team has also expanded beyond 20 employees.

Partnership-focused industrial strategy

IPercept describes other industrial monitoring companies as potential partners rather than simply competitors.

There are several companies addressing adjacent problems in industrial equipment monitoring, but their technical approaches differ.

Augury uses vibration and ultrasound sensors to monitor general rotating equipment.

Samotics analyzes electrical signatures to identify equipment conditions.

MachineMetrics focuses on CNC monitoring through connections to machine controllers.

IPercept's approach instead focuses on measuring the physical motion of CNC equipment without requiring controller access.

The different approaches mean that industrial monitoring can involve multiple layers of sensing and diagnostics depending on the machinery and application.

For IPercept, partnerships with industrial service providers are an important part of its expansion strategy.

The company intends to strengthen those relationships using the latest funding.

$16.5M to accelerate US expansion

The Series A follows IPercept's previous €5 million seed round, which closed in May 2025 and was also led by Luminar Ventures.

The new $16.5 million financing provides significantly more capital for the company's next phase.

One of the immediate priorities will be entering the United States.

IPercept plans to make its first dedicated US hires as part of the expansion.

The company also plans to strengthen its partnerships with industrial service companies that can integrate its monitoring technology into existing maintenance agreements.

For a Swedish industrial technology startup, the US represents an important potential market given the scale of its manufacturing base and the presence of major automotive, aerospace, energy and industrial companies.

Physical AI moves deeper into factories

Much of the current attention around physical AI has focused on humanoid robots, autonomous vehicles and warehouse automation.

IPercept is applying the same broader idea of AI interacting with the physical world to a less visible but critical part of the industrial economy: CNC machines.

Its technology combines physical sensing, machine diagnostics and AI-driven analysis to understand equipment without requiring direct access to machine controllers.

The company now has more than 30 enterprise customers and is expanding beyond Sweden.

With $16.5 million in new capital, IPercept is preparing to scale its technology, deepen industrial partnerships and establish a presence in the US.

Its long-term goal, according to Szipka, is to help Western manufacturers increase speed and competitiveness by making industrial equipment more intelligent and easier to maintain.