Digital Energy Journal- Emerson - using local AI models in automation

Join | Existing Member Login

You are Home » News » View Article

Emerson - using local AI models in automation

Thursday, December 11, 2025

Automation technology company Emerson is developing 'local' (ie not cloud based) AI models to help manage industrial processes, including designing plant, plan control system upgrades, and develop decarbonisation strategy

The AI systems which are most useful in industrial facilities and control systems may be models running locally, not large language models (LLMs) in the cloud, says automation technology company Emerson.

Where an LLM is trained on all the text data in the internet, and works mainly by finding patterns in language, a 'local model' can work with data using physics and engineering equations, and traditional AI.

These 'local models' can be specifically built for the process they are looking after, and able to apply the relevant physics and engineering equations.

They can also use calculations based on first principles (direct equations), rather than just looking for patterns, as a LLM does.

Because they have been built and tested for a specific piece of equipment, their results should be more reliable than those from a broad large language model based system, Emerson says.

It also means that all the processing can be done on local machines, without using cloud systems.

'Generic large language models are insufficient for a live, mission-critical industrial plant that cannot afford unreliable or unsafe results,' said Ram Krishnan, chief operating officer at Emerson.

One example is Emerson's AspenTech 'Optiplant AI Equipment Layout' tool, which uses generative AI to generate multiple design options for greenfield and brownfield plants.

Engineers can then select the layout which they think is most suitable, taking into account performance, buffer zones, safety, proximity, and other factors.

Another local AI tool is 'DeltaV Revamp', which works out how to evolve a legacy control and safety system to using the modern DeltaV communications protocol.

The AI 'learns' from data gathered from thousands of modernisation projects, so it can plan the control system update with increased speed, and accuracy.

AspenTech's 'Strategic Planning for Sustainability Pathways' local AI tool uses generative AI technology to help companies work out the right long term decarbonisation strategy, taking various options and possible pathways into account.

Emerson has its 'Aspen Virtual Advisor' (AVA) built into its technologies. This provides users with operational guidance when they are working with its tools. AVA is programmed with deep knowledge of the relevant software tools.

Users can ask questions in natural language, to identify possible bottlenecks in processes, or find better ways to achieve specific goals.

Emerson has virtual advisors built into its Ovation 4.0 Automation Platform, which can be used to run 'software-defined automation', and its Guardian Digital Platform for managing product support, subscriptions and licenses.

At the company's 'Emerson Exchange' conference in May 20 2025 in San Antonio, Texas, it announced its new platform 'Project Beyond' which can be used to deploy and manage its AI applications and models.

Associated Companies
» AspenTech
» Emerson

CREATE A FREE MEMBERSHIP

To attend our free events, receive our newsletter, and receive the free colour Digital Energy Journal.