A Smarter Way to Monitor Your Equipment with IBM's Asset Monitoring and prenode Edge AI

Many software providers, including IBM, want to enrich their products with machine learning (ML) to deliver AI-based digital services that meet the evolving needs of businesses and customers in today's digital landscape.

However, ML needs enormous amounts of data that cannot easily be accessed because customers are unwilling to share their data due to privacy or high data volumes. In addition, companies often face the challenge of data being distributed across silos, making it difficult to use all the data for ML effectively. It causes knowledge of silos and delayed forecasts.

IBM actively pursued a technical solution to overcome data challenges and enhance their IBM Maximo® Monitor portfolio (a remote asset monitoring application) with AI capabilities.

To resolve these challenges, our decentralized AI solution, prenode mlx, offers a comprehensive infrastructure for on-device machine learning, enabling the training of models on federated and isolated data sets. This solution allows training ML models without sharing data with a central entity and keeps the data on edge at individual sites to ensure privacy, security, and accurate predictions.

In collaboration with IBM, we developed a Hammer-Rig showcase where mlx seamlessly integrated into IBM's infrastructure, demonstrating its capability to predict overheating events by analyzing data from multiple pumps. Our decentralized AI software analyzed sensor data from each individual rig, accurately predicting overheating events and sending the predictions to IBM Maximo Monitor platform through IBM Cloud. This enables machine operators to quickly receive warnings when an overheating event is predicted.