Projects

Projects by Eberle in practice: robot palletising, AI cheese inspection, retrofit, process monitoring with control, storage and twin.

Eberle designs, builds and automates machines and systems for industry, food production and energy supply. A selection of projects shows how robotics, image processing, control technology and our software work together in operation.

Playing the video loads it from YouTube (Google), which transmits your IP address. More in our privacy policy. Watch on YouTube

Cheese wheels in a care machine, below camera images of wheels inspected as good and with detected defects

Industrial image processing

AI cheese inspection at Gebr. Baldauf

Cheese wheels often ripen for many months. Checking thousands of them regularly for mould is impossible by hand. Together with Leu-Anlagenbau, Eberle therefore developed an automated inspection for the Allgäu cheese dairy Gebr. Baldauf.

  • A 4K colour line-scan camera on the mobile care robot records every wheel before it is treated.
  • MVTec HALCON analyses the images with deep learning and detects defects such as mould or cracks.
  • The results go to storage: every wheel is fully documented and traceable.

The system has been in operation since December 2024. The full article: AI meets cheese production: rethinking efficiency. More under industrial image processing.

Temperature curves of a blood refrigerator over several days in the control browser view

Process monitoring with control

Data at a glance, across sites

With control, measured values, alarms and operating data from distributed systems come together in one place. Limit values trigger notifications by text message, e-mail or phone call, and the recording also serves as documentation.

Long glazed production line by K&S Anlagenbau in a factory hall

Simulation with twin

Testing machines before they are built

twin creates a digital twin of the machine. Control programs are tested on it before the first part is manufactured.

  • K&S Anlagenbau: twin validates production lines that process thousands of parts per minute as early as the concept phase.
  • igm Robotersysteme: the welding robot manufacturer saves around 40% of the time with virtual commissioning and a 3D user interface.

Detailed case studies on control, storage and twin are available on digifai.com.

Importing a robot and simulating the sequence in twin

Planning a similar project?

Tell us about your task. We will tell you how we would solve it.

Get in touch