The Berlin startup points cameras at production lines, converts the footage into live data, and says it has lifted output at Bosch and ABB plants. Now it wants the US.

Berlin startup that films what happens on a factory floor and turns it into live production data has raised €16.3M in Series A funding.

Almetra, until recently known as Deltia, mounts AI cameras above assembly lines at companies including Bosch, Siemens Energy and ABB, then analyses each work step to find where output is being lost.

According to the company, customers have raised productivity by roughly 20% as a result.

The round was led by blisce/, the transatlantic fund behind Spotify and Pinterest, with participation from NAP, Merantix Capital, Robin Capital, Underline, Critical Ventures and a group of business angels.

Almetra’s own customer figures are more specific: eBike Systems, a Bosch subsidiary, reported a 19% increase in output within weeks of deployment, and ABB reported a 15% improvement in productivity.

Read together, the round 20% reads as a convenient average of plant-level results that vary by site rather than a single measured number.

Almetra was founded in 2022 by Maximilian Fischer and Silviu Homoceanu as a venture built out of Merantix, the Berlin AI studio whose first fund also incubated names such as Vara and Cambrium.

It rebranded from Deltia this year, and the new name marks a wider shift in strategy. Where Deltia analysed manual production steps, Almetra is positioning itself as a broader “data and control layer” for the factory, a single source of operational intelligence that pulls together video, machine data, IT systems and what operators know.

The mechanics are deliberately light-touch. Cameras process video locally and convert footage into structured data, such as cycle times, output rates and equipment utilisation, without requiring integration into a plant’s IT systems.

Worker privacy, a live question whenever cameras go up over people doing their jobs, is built in: footage is anonymised, most of it never leaves the site, and only short, randomised snippets are kept for root-cause analysis.

The pitch is that a camera is the cheapest and fastest way into a factory, and that once inside, it can become the foundation for everything else.

“Plant teams know they are losing capacity, but not where or why. We give them certainty instead of guesswork,” said Fischer, who serves as chief executive. He said most customers find significant optimisation opportunities within the first few weeks.

The no-integration approach is what Almetra uses to distinguish itself from the heavier names in industrial AI. Sight Machine, which has raised more than $85M, requires deeper embedding and targets larger enterprise deployments; Augury focuses on machine-health monitoring through vibration and ultrasound sensors.

The global machine-vision market was valued at about $20.4bn in 2024 and is projected to reach $41.7bn by 2030, a market Almetra is far from alone in chasing.

That argument has drawn attention beyond industrial software. Almetra has been accepted into Google DeepMind’s Robotics Accelerator and a Physical AI Fellowship run by AWS, Nvidia and MassRobotics, recognition that a company sitting on continuous video from live factory floors is a potential data source for the next generation of industrial robots, not just a productivity tool.

“Manufacturing is the backbone of the global economy,” said Sam Giber, general partner at blisce/, adding that Almetra had shown it could deliver “measurable results” for leading industrial firms.

The company, which now employs around 40 people, will use the capital to expand into the US and to keep building out the platform, including robotics applications in selected production environments.