Consider a technician arriving at a substation to investigate a circuit breaker reporting an operating problem. Before deciding what needs to be done, the technician must understand the asset’s maintenance history, previous issues and any technical documentation relevant to the job.

That information is likely to exist, but finding it can take time.

The technician may need to search different applications, access multiple folders and review historical records before assembling enough information to understand the problem. For utility field crews, particularly those working across transmission and distribution networks, this is a familiar challenge.

I see one of the most valuable applications of GenAI in making that information available through a single digital assistant on the tablet a field worker already carries.

The principle is straightforward. Instead of searching through several systems, the technician asks a question and receives an answer drawn from the relevant documentation. The technology is useful because it simplifies an existing task and reduces the time spent retrieving information.

Bringing technical knowledge into one application

Utility field workers typically carry a tablet alongside other specialist equipment. These devices already provide access to different applications and technical records, but accessing information across them can be time-consuming.

Take the circuit breaker example. A technician could scan the asset and ask a digital assistant to show its maintenance history, recorded problems and relevant documentation.

Without GenAI, answering that question might involve navigating multiple portals, databases and folders, then manually reading and comparing the information.

A properly configured GenAI application could retrieve relevant records from those sources and present the information through one interface.

This is where large language models and small language models become useful. When connected to the appropriate knowledge sources, they can help interpret a field worker's question and retrieve information relevant to the task.

The important difference is the speed at which the information becomes available.

A field worker who can access the required context through one application spends less time searching for it. That makes the process simpler and allows the technician to concentrate on understanding the issue at hand.

For me, that is the immediate opportunity: one application instead of many.

Reducing the effort involved in everyday fieldwork

The same principle applies to the administrative work surrounding field operations.

Manual effort often comes from having to locate information across different systems. Each application may contain part of what a worker needs, but the individual still has to navigate between them to build a complete picture.

Bringing that information into a single interface reduces the number of steps involved.

It also makes GenAI useful without requiring utilities to change every aspect of how their field teams operate. Workers can continue using familiar devices while gaining a more direct way to access the documentation they need.

The objective is to simplify information retrieval, particularly when workers are dealing with technical issues that require them to understand previous activity and the condition of an asset.

Transferring experience to the next generation of workers

Another important application concerns knowledge transfer.

Long-serving utility employees accumulate considerable experience working with particular assets, environments and systems. Some of that knowledge is captured in technical documentation, maintenance records and other materials built up over time.

When a new employee joins, there are generally two ways that knowledge can be transferred.

The first is person-to-person interaction. A new worker sits with an experienced colleague, asks questions, takes notes and gradually builds an understanding of the work.

The second is system-to-person interaction, where the employee learns by accessing information already captured in the organization's systems.

GenAI can make the second approach considerably easier.

If existing knowledge documentation is connected to a digital assistant, a new field worker can ask questions directly and receive information drawn from those materials.

That could reduce the time spent locating and working through documentation during onboarding. It also creates a practical way to make accumulated technical knowledge more accessible to employees who are still becoming familiar with their environment.

The value depends on having that knowledge captured and available in the first place. A digital assistant needs reliable documentation to retrieve from, which makes the quality and accessibility of existing information important.

Why utilities need a deliberate approach

The potential productivity benefits are considerable, but utility companies operate differently from many other industries.

Utilities are highly regulated and understandably cautious about adopting new technologies. Security is a major consideration, particularly when new applications need to access operational information.

I consider that caution a strength.

In a fast-moving technology business, an AI application might be introduced and refined relatively quickly. For a utility, adoption needs to be more deliberate because of the environment in which the technology operates.

That is why I believe implementation should begin with a purpose-driven use case and a clearly defined user persona.

A field technician, for example, has specific information requirements. The GenAI application should be designed around those requirements and the work that person needs to perform.

Human involvement also remains essential.

There are field activities involving isolation, testing, verification and authorization where a person must remain part of the process. GenAI can help retrieve information and support the worker, but it should operate within the governance requirements surrounding those activities.

Data integrity, security and governance are equally important. Utilities need to consider where information is stored, how it is retrieved and which users or applications are permitted to access it.

These decisions should be made before introducing GenAI into field operations.

Preparing for the years ahead

I expect GenAI adoption in utilities to grow, although the pace will reflect the industry's emphasis on security and operational control.

That measured approach is particularly relevant as utilities prepare for increasing demands on their infrastructure.

The State of Texas provides an example. Significant load growth is expected around 2030 and 2031. Over the next three or four years, I believe GenAI could transform how utilities operate and potentially accelerate the development of transmission lines, distribution networks and substations.

For field workers, the benefits begin with practical improvements: finding technical information faster, reducing the effort involved in searching multiple systems and making experienced employees' knowledge more readily available.

The technology needs to be introduced with a clear purpose, the right data controls and appropriate human involvement.

Utilities are right to be thoughtful about adoption. In an industry responsible for critical infrastructure, slow and steady progress can go a long way.