| Monitoring & inspection
AI agents move into operations
Emerson is extending its Ovation automation platform with AI agents designed to support plant diagnostics, maintenance and control, as artificial intelligence moves closer to day-to-day power and water operations
ARTIFICIAL INTELLIGENCE IS increasingly being incorporated into systems used to operate and maintain power and water infrastructure. Emerson’s latest development aims to embed AI directly within established operational workflows. Emerson is expanding its Ovation Automation Platform 4.0 with five new Ovation AI Agents covering control sequences, alarms, predictive maintenance, root cause analysis and control loop performance. The agents run continuously, analysing plant and equipment conditions and providing recommendations when potential problems or opportunities for optimisation are identified. For hydropower operators, the approach reflects challenges including more flexible plant operation, ageing equipment and growing volumes of operational data. Emerson says the agents can reduce diagnostic time by correlating information that might otherwise require operators or engineers to investigate several systems or datasets manually. Unlike chatbot-style AI interfaces, the agents are
integrated with Ovation 4.0 and its existing control, alarm and maintenance workflows. They are also time synchronised with the Ovation system. “Ovation AI Agents deliver a major step change in our AI portfolio, representing a significant advancement in how we help our customers build more intelligent, efficient and resilient operations,” said Bob Yeager, president of Emerson’s power and water solutions business. “The AI-enabled agents handle complex data analysis and pattern recognition to monitor asset health, run simulations, forecast failures and prioritise interventions. These powerful tools enable operators to shift from reactive to proactive planning without disrupting daily workflows.”
Five areas of focus The Sequence Assistant Agent analyses real-time
deviations and previous failures to provide guidance when automated sequences do not proceed as expected. Emerson gives the example of a steam turbine start-up that stalls because a required permissive condition has not been met. Rather than manually examining logic diagrams and point values, the agent can identify the failed condition and recommend corrective action. Although the example comes from thermal generation, the approach could have applications across generating plant using automated sequences. The Alarm Insight Agent analyses and prioritises alarms during an alarm flood, helping operators distinguish significant events from secondary alarms accompanying a disturbance. The Predictive Maintenance Agent uses equipment condition data to detect degradation and provide early warning of potential failures. The Root
Cause Agent correlates alarms and trips with live process and equipment conditions to identify underlying causes. The Loop Performance Monitor Agent continuously assesses control loops for problems including oscillation, deadband and poor tuning, generating recommendations for engineering review.
Keeping operators in control Emerson says the agents have been developed
around a human-in-the-loop model. They can monitor conditions, conduct analysis and recommend actions autonomously, but operational decisions remain with plant personnel. Recommendations are presented for approval rather than automatically implemented. According to Emerson, early deployments indicate the approach can reduce diagnostic times. In one case, the Sequence Assistant Agent identified a failed permissive condition in seconds, compared with what the company says could previously have required hours of investigation. Emerson also reports that a Predictive Maintenance Agent used by a major utility identified an equipment anomaly before it became critical, allowing action before an unplanned outage. The company has not disclosed the utility or equipment involved. The examples illustrate a potential role for AI in power generation: reducing the time specialists spend finding and correlating information before making decisions.
Expanding the agent portfolio The first five Ovation AI Agents are scheduled to become
available this autumn, with customers able to deploy individual agents according to their requirements. Emerson says another 28 agents are on its roadmap, covering operations, maintenance, optimisation, instrumentation and control, management and other operational areas. It is also developing an Ovation AI Agent Builder to allow users to design and customise agents for specific applications.
www.waterpowermagazine.com | September 2026 | 27
Designed to seamlessly integrate with the Ovation Automation Platform 4.0 release, the new Ovation AI Agents are embedded into native operational workflows and operator tools, eliminating the time-consuming need to switch between systems
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45
