Secure, local AI plant monitoring without reliance on the cloud – sign up for our newsletter to get the latest updates!
Secure, local AI plant monitoring without reliance on the cloud – sign up for our newsletter to get the latest updates!
We minimise unplanned downtime and service call-outs in industrial refrigeration systems through data-driven early diagnosis.
On-premises AI or cloud-based AI: what’s the difference?
The difference lies in where the AI runs and where the data has to travel for it.
Cloud AI: Operational data is sent over the internet to an external data centre, analysed there, and the result is sent back. This allows for very large models, but it requires an internet connection, and the data leaves the company.
Local AI (Referee AI): The model runs on a device directly on site – in the case of Referee AI, on the hardware module in the control cabinet. The data stays within the company network, the analysis works without the internet and does not depend on any external service.
Six reasons why your refrigeration system runs more efficiently with our local AI.
Referee AI monitors your systems round the clock. Your technicians are alerted only when necessary, rather than wasting time on routine checks.
The team comprises specialists in plant engineering, refrigeration technology and AI, with over 20 years’ experience in the refrigeration industry and 15 years’ experience in predictive maintenance.
The hardware module is connected directly to the existing control system without the need to replace any existing equipment. In compatible systems, it can be installed in a matter of hours.
The AI analyses all the data locally within a closed system. Without the cloud and without an internet connection, there is no external access.
We develop the module, AI models and user interface ourselves. From installation through to after-sales support, you’ll have a single point of contact who is familiar with the entire system.
Referee AI does not rely on data centres or external cloud providers. The system operates independently on-premises, and you retain control over your data.
Plan maintenance rather than reacting to breakdowns. Traditional plant monitoring often only reacts once thresholds have been exceeded or a fault has already occurred. Referee AI goes one step further: the system continuously analyses the interplay of measured values and detects gradual changes in plant behaviour before they result in unplanned downtime. This is all the more important as more and more plants are being managed by fewer and fewer technicians: service call-outs can be planned in a more targeted and efficient manner.
Sensor data is turned into specific recommendations for action. Your team knows immediately what to do.
A virtual model of your plant reflects its operational status in real time, for analysis, simulation and predictive maintenance.
The hardware module connects directly to your existing PLC without the need for additional cabling. It continuously records pressure, temperature and other operational data in real time.
All data is clearly displayed in a central user interface on your company network. Live monitoring and a digital twin show the status of your plant at a glance.
Anomaly detection runs directly on the module at the plant. It detects deviations and trends before minor signs develop into costly faults.
The AI diagnosis states the probable cause in plain language and provides clear recommendations for action. Everything runs entirely locally: no cloud, no internet connection required, full data sovereignty.
