01 / PREDICTIVE
02 / AI DRIVEN
03 / INDEPENDENT
The unexpected failure of an electrical engine can break your production and cause high loses. APEMAN platform predicts upcoming failures to give you needed time for the assets maintenance while preventing loses from manufacturing disruptions
APEMAN Edge and Server AI features uses neural networks for anomaly detection to prevent failures.
Device trained modules are continuously improved to deliver superb failure sensitivity without user interaction or device dedicated setup.
Embedded AI is as close to the source of streaming data as possible thanks to dedicated monitoring edge devices.
Devices are independent from on-site infrastructure and capable to process data offline while using cloud storage when possible.
The APEMAN is a distributed edge-cloud Digital Twin model applicable to predictive maintenance of machines on the shop floor. An edge device – Integrated Sensing and Analysis Box (ISAB) is able to collect and analyses data using deep neural network models in order to monitor the machines and predict their failures.
For the production companies in which unexpected electric engine failures causes significant loses APEMAN is predictive maintenance system informing in advance about upcoming failures.
Comparing to competitors APEMAN is able to predict failures that were observed and sensor values describing them are unknown.
The system is based on the Apache tool chain of the MIDIH architecture and uses both levels of data processing – the Data-in-Motion is responsible for real-time analysis of incoming data directly on the edge devices, whereas the Data-at-Rest layer is used to validate and retrain the real-time models.