Viessmann x Almetra

Viessmann · Legnica plant, Poland

Before Almetra, employees at Viessmann’s Legnica plant had to record by hand how long each individual work step took. Today, cameras and AI capture that data continuously and objectively - and productivity on the equipped lines is up 20%.

THE CHALLENGE

Viessmann, the German heating technology group, produces heat pumps and heating systems at its plant in Legnica, Poland. With demand for heat pumps surging across Europe, the plant was under pressure to get more output from its existing manual assembly lines - without simply adding headcount in a tight labour market.

To raise productivity, the team needed to understand how long each individual production step really took, where time was lost, and which stations were holding the line back. But the data they had to work with was thin and unreliable.

WHY THE OLD WAY WASN’T ENOUGH

Manual recording by operators

Workers had to log by hand how long each work step took. Entries were incomplete, inconsistent, and took time away from actual assembly work.

Snapshots instead of a continuous picture

Time studies captured a few cycles on a good day - not the everyday reality of short stops, waiting times and extra motion between stations.

Output known, process unknown

Production systems showed what came off the line, but not how it was produced. The root causes of lost time stayed invisible.

The team had plenty of improvement ideas - what they lacked was objective evidence to prioritise them.

ALMETRA’S APPROACH

Almetra connected the manual assembly areas in Legnica to standard industrial cameras - covering the work of 45 employees - and used AI to turn the video feeds into continuous process data. No special cameras, complex calibration or dedicated lighting were required.

Instead of a stopwatch and a clipboard, the plant now had an automatic, continuous record of every cycle: how long each step took, how often operators had to walk or wait, and where cycle times deviated from the standard.

How the team used AI to see more

Automatic cycle and step times

Every work step was timed automatically, replacing the paper-based recording that employees had previously done themselves.

Walking routes and waiting time made visible

The analysis revealed unnecessary motion and idle pockets between stations - the small losses that never showed up in reports.

Video evidence for every finding

Each deviation was linked to a short, anonymised video clip, so engineers and operators could review the root cause together and agree on the fix. Raw footage is deleted; only anonymised snippets are retained.

Best practices for training

The fastest repeatable cycles were identified and used as reference for onboarding and retraining operators.

THE RESULTS

20% productivity increase

Achieved on the lines equipped with Almetra since installation at the Legnica plant.

Up to 50% expected plant-wide

Plant manager Bartłomiej Magiera expects productivity to rise by up to 50% once the technology is rolled out across all of the plant’s lines.

Manual time tracking eliminated

Operators no longer record step times by hand - the AI captures everything automatically and objectively.

Data-driven line balancing

With real step times for every station, the team can rebalance work, remove waiting time and plan staffing based on evidence rather than estimates.

Ready to see what AI could reveal on your line?