Information from the current-time profile replaces the sensor
Anomalies in load behavior leave clear “fingerprints” in the current-time profile. Whether it’s a worn bearing, a broken fan blade, air in a pump or whatever: If the machine or system is not running “smoothly”, characteristic changes occur over time in the power consumption.
Elec-Con detects these changes compared to normal operation with the help of AI and raises the alarm – often a week or two before damage or failures become apparent.
The alarm message can be transmitted to higher-level controls via MQTT – depending on the customer’s wishes – or as an SMS to one or more smartphones using additional peripherals.
Don’t be afraid of AI and the corresponding programming: Elec-Con has gained a lot of experience in several research projects with the Deggendorf University of Applied Sciences and makes this available to its customers.
The knowledge of the power supply has so far hardly been used to detect anomalies in the connected load or to obtain and evaluate operating data. However, since the data in the power supply is available with high temporal resolution, valuable information about the system status can be obtained from it. The scale ranges from simple wear detection, to the detection of anomalies and the detection of intruders, to the detection of intruders and the detection of anomalies. Elec-Con even accommodates simple AI models on the customer-specific power supplies of the diPSU series; the power supply thus detects anomalies in the load independently – without an edge computer.
