For the determination of the battery condition, our artificial intelligence needs at least information about

• Power
• Voltage
• and temperature

in a resolution of 1–60 seconds. Customers who manufacture or operate battery storage systems can usually provide us with this data.

Battery diagnostics without prior information

But if not, that’s not a problem either. If we have no or too little data on batteries, we use our patented hardware to test the frequency-dependent internal resistance in the batteries or battery cells.

Within a certain frequency range, current is sent into the battery, which generates various response signals. This gives us a non-invasive image of the chemical processes inside the battery, which our AI can then evaluate in seconds.

Quality of initial values decisive

Any Artificial Intelligence relies on reliable output values to provide correct results. Data gaps and errors in data collection (e.g. due to defective sensors) can lead to errors that affect forecast accuracy. At NOVUM, we not only have a sentinel AI that constantly outputs how reliable our analyses are, we also take a look at your data basis together with you and also compensate for possible data gaps if necessary. This means you always get statements at the highest level — regardless of the data basis.