Over 10,000 lineworker jobs will remain unfilled each year through 2033, according to the U.S. Bureau of Labor Statistics. But some of the power-hungry technologies driving labor-intensive grid upgrades also can help utilities meet that demand.
A prime example is artificial intelligence (AI), whose data centers require unprecedented amounts of electricity to run and cool their servers, routers, and other equipment. That’s why Microsoft recently signed a 20-year contract to buy all of the electricity from the Three Mile Island Unit 1 nuclear reactor, which is capable of generating about 837 MW. But it would take six of those reactors to power just one of the 5 GW data centers that Microsoft and other AI companies plan to build.
AI is key for ensuring that the grid can meet that demand, from generation through transmission and distribution. One example is the solar farms that AI companies such as Google are increasingly using to power their data centers. Drones equipped with computer vision — a form of AI — can thoroughly inspect both the solar panels and the transmission lines more efficiently than dozens of people whose skills are already in short supply. Drones also help keep those inspectors safe by eliminating the need to work at height. And when lineworkers do need to climb a pole or tower, the AI enables them to work more safely and efficiently by pinpointing exactly where the repairs need to be made.
Preventative maintenance is another use case. By analyzing photos taken by drones or satellites, AI can flag areas where vegetation is starting to encroach on transmission and distribution lines. This minimizes the need for — and expense of — manually inspecting lines by directing crews to only the sections that need clearing. Another example is using AI to analyze those photos for signs of overvoltage damage due to lightning strikes or overheating.
AI also helps utilities with digital transformation initiatives that increase reliability and efficiency. One big challenge is that the process of digitizing a century’s worth of analog grid records requires extensive manual review because the data often is inconsistent or incomplete. Southern California Edison (SCE) estimates that a manual review of its 1.4 million-plus power poles would require 300,000 worker hours over multiple years and cost $16 million.
AI streamlines that task by analyzing and verifying each piece of photo data, including identifying the location of each asset within 30 feet so crews don’t waste time looking for that equipment. Those insights also reduce restoration times and can minimize or avoid fires by enabling lineworkers to go straight to the location of downed infrastructure.
“The tool substantially reduces the need for manual processes to evaluate millions of photos and provides reliable data that is easily searchable,” says an Edison Electric Institute case study of SCE’s digital transformation. “By incorporating the data more efficiently and using an automated process, the company is expected to save $8 million and 170,000 worker hours across SCE’s service area.”
The Right Mobile PCs Unlock AI’s Full Potential
When developing an AI strategy, utilities shouldn’t overlook the role of the laptops and tablets that lineworkers and other employees use. For example, when those devices have embedded 5G cellular modems, crews can quickly and securely access grid information stored in cloud data centers. They also can upload photos, maintenance records and other data to those data centers for storage and AI analysis. By choosing mobile PCs with embedded cellular, utilities can leverage the private 5G networks that they already own to support applications such as AMI/AMR and drone inspections.
AI is far more compute intensive than other applications, which is why AI data centers require more power for operating and cooling. The same applies to laptops and tablets — with one crucial difference: Data centers are connected to the grid, while mobile PCs rely on batteries. Hence the importance of choosing devices that balance compute performance with battery life. Look for a battery life of at least 10 hours and ideally up to 20 with an optional high-capacity battery. This ensures that the tablets and laptops can last not only an entire normal workday, but also the extended shifts that are common in the aftermath of storms.
A passive cooling system is an ideal way to extend both battery life and device life in tablets and laptops. Eliminating fans frees up battery power for the processor, display, and other components. It also extends device life by eliminating moving parts that can eventually wear out.
Some AI applications need to run on the device rather than in the cloud. Look for specs such as Intel® Core™ Ultra 5 and 7 processors with Intel vPro®, an integrated neural processing unit (NPU), and options such as an NVIDIA RTX™ A500 professional GPU. These ensure that the device can handle even the most compute-intensive applications and support generative AI virtual assistants.
The right mobile PCs enable utilities to leverage AI to work smarter rather than harder. That efficiency and productivity is key for overcoming the chronic shortage of lineworkers and other skilled employees while maximizing grid reliability.
