IBM is integrating artificial intelligence agents into its Maximo Application Suite to revolutionize industrial maintenance. The shift is from fixed-interval servicing to a smarter, condition-based system driven by real-time machine data. Much like maintaining a vehicle based on its actual condition rather than a manufacturer’s schedule, this approach aims to cut costs, extend asset lifespan, and prevent breakdowns.
For nearly 40 years, Maximo has supported asset lifecycle management. Today, with the help of sensors and visual inspection technologies, it’s now possible to continuously monitor the actual condition of equipment.
Today, IBM is going even further with the addition of three AI agents to the platform:
– Maximo Assistant – A conversational agent powered by a large language model (LLM) that allows users to query data easily through a chat interface, without requiring technical skills.
– An Optimizer (coming July 2025) – Recommends the optimal time to replace equipment by considering performance metrics and budget constraints.
– Condition Insights Agent (expected end of 2025) – Automatically collects asset data to estimate current condition and predict potential failures.
These innovations are accelerating the shift toward predictive maintenance—more efficient, cost-effective, and sustainable.
Case Study: Smart Monitoring of an Industrial Chiller
At IBM, the Global Real Estate (GRE) division manages over 115,000 industrial assets. Through its intelligent buildings project powered by Maximo, the company has already saved $1 million annually in energy costs.
Now, a new pilot project is testing the Condition Insights Agent on a $2 million industrial chiller located in Poughkeepsie. This type of machine includes rotating components such as ball bearings, which can exhibit early signs of wear through micro-vibrations undetectable by humans. With AI, such anomalies are caught early—enabling planned preventive maintenance and avoiding costly breakdowns.
Maximo consolidates all available data—asset model, expected lifespan, maintenance history, sensor data—to provide a comprehensive, real-time view of asset health.
According to IBM researchers, even a slight drop in efficiency or an increase in energy consumption can be an early sign of failure—automatically detected by the agent.
Advanced Analysis Through Time-Series Modeling
To enhance these capabilities, IBM plans to integrate time-series analysis models into Maximo. This will allow the Condition Insights Agent to identify trends and anomalies that are invisible to the human eye within the sensor data streams—further improving diagnostic accuracy.
An AI-Orchestrated Vision
Maximo Assistant, which will soon be updated with Granite 4.0, will act as a conductor, orchestrating these specialized AI agents. A third agent, expected in 2026, will help plan long-term asset investments. Going beyond Maximo’s current optimizer, it will enable users to define specific investment criteria—such as operating costs, budget constraints, and sustainability goals—to guide asset replacement decisions.
To support this vision, IBM has launched AssetOpsBench: the first agent-oriented test and benchmarking environment for asset management. It provides the community with a platform to test, validate, and improve AI agents for smarter industrial asset management.
Reference : https://research.ibm.com/blog/Maximo-automating-asset-management
