Generative AI for Edge-Based Predictive Maintenance in Smart Factories

Abstract

In this article, we introduce a lightweight AI system named GenAI, predestined for industrial IoT devices to enable proactive predictive maintenance at the edge. The proposed technique for monitoring and generating synthetic sensor data tries to rebuild the meaningful signal variation in the small common datasets and transfer it into sensor signals that represent the actual data variability regarding equipment operation. Reducing downtime to avoid maintenance requirements is especially critical in advanced factory settings, where equipment failures are too costly. Furthermore, the local processing of generative bidding at the edge does not need to be connected to the main network, meaning that data must be protected from neighbor interference. Through our new platform strategy, the apparent issue of the acute underdevelopment of local data within a distributed and hugely scalable system is explained in order to prove its applicability within big industrial applications.