References
Projects at the interface of machine, automation and data.
A selection of completed projects, from model-based optimisation and AI condition monitoring to the complete, production-ready special machine.
Winner of the Wood Energy Award 2025 · Category Pellets. Nominated for the RIZ UP GENIUS 2026
Solarpellet®: AI-optimised pellet production powered by photovoltaic energy
Solarpellet combines mechanical engineering, automation and artificial intelligence in a completely newly developed production plant. At its heart is an AI algorithm developed by Zellhofer Engineering that adapts production in advance to the available photovoltaic energy.
To do this, weather forecasts, expected PV output and current machine states are brought together. Based on a digital model of the plant, the system plans autonomously when and at what output the individual production steps are run. The aim is to use as much locally generated solar power as possible directly, to avoid load peaks and at the same time to ensure a stable and efficient production process.
Zellhofer Engineering was responsible for the entire development, from the mechanical plant concept through automation and control technology to AI-based operational optimisation and the production-ready machine.
AI-based production planning
Digital twin
Machine learning
Mechanical engineering & automation
Forecasting & optimisation
Forecasting and optimisation tool for a renewable energy community
For a renewable energy community, a system for forecasting PV generation and consumption was developed. Based on historical measurements and external influencing factors, generation and load are predicted and used to optimise the energy flows.
The aim is to use the locally generated PV power within the energy community as efficiently as possible.
Forecasting models
Load analysis
Energy flow optimisation
Predictive maintenance · IoT · Retrofit
AI condition monitoring and remote monitoring for pump units
For existing pump units, an IoT-based system for continuous condition monitoring and remote monitoring was developed and implemented.
Operating data is recorded continuously, made available centrally and evaluated using AI-based anomaly detection. This allows deviations from normal operating behaviour to be detected early and plant conditions to be monitored remotely.
The solution was integrated into an existing plant as a retrofit.
IoT
Remote monitoring
Anomaly detection
Predictive maintenance
Retrofit
