Mud-dled Thinking: The Impact of Artificial Intelligence on Today’s Municipal Utility District

Municipal utility districts, or MUDs, might not sound glamorous, but they’re the infrastructure lifelines of countless communities. They allow for infrastructure construction, maintenance, and improvements in places not served by a city or other municipal utility. Like many local entities, they are burdened by paperwork and procedural muck, which are painstaking and time-draining for humans to work through.

Like the rest of the world, MUDs are beginning to dig into something new: artificial intelligence. AI isn’t just a buzzword in big tech. At the neighborhood level, it’s making public service quicker and more efficient.

Unincorporated areas can find MUDs extremely useful in funding initial infrastructure, says Jeff Peña, P.E., Senior Project Manager for Pape-Dawson, “but 15 years after their formation, things aren’t shiny and new anymore. They often need major rehab. That takes planning, and that’s where we come in.”

Pape-Dawson’s work with MUDs involves a cross section of drainage, transportation, water facilities, wastewater facilities, linework for water and sewer, storm sewer, channel rehab, and more—and AI is making that work more efficient. Whether AI is driving a fourth industrial revolution, as some claim, is an open question. What’s certain for Jeff and his team is that their work is already seeing marked productivity gains. Pape-Dawson’s experience so far largely matches the results of a Harvard Business Review study, which found AI use resulted in better task completion rates, faster completion times, and higher-quality results.

The Pape-Dawson crew started out the way many of us have, by simply trying it out and getting comfortable with its capabilities. “At the beginning, we found AI can be really good at brainstorming and coming up with questions,” says Jeff. Since then, his team has expanded their use of AI.

Jeff couches his enthusiasm with skepticism, though, and insists on employing AI as a tool, not a crutch. “As engineers and consultants, we’re responsible for our product,” he cautions. Validating outputs and ensuring high-quality deliverables are still the purview of humans, so flesh-and-blood engineers should oversee every aspect of a job for a MUD or any other client.

On that point, Jeff has more to say: AI can hallucinate, and conscientious engineers must actively guard against flat errors as well as more-subtle biases. “We’ve scanned handwritten notes from an operator into our enterprise AI, and the AI misread some of the values,” says Jeff. Without a human looking over the AI’s virtual shoulder, bad calculations could result.

Vice President Jason Johnson, P.E., agrees: “It can save loads of time, but you’ve got to check it.” On at least one occasion, Jason has redone the AI’s calculations and come up with a different answer. “When I confronted the AI about it, the AI admitted I was right,” he says with a grin.

Still, time savings can be significant. “You can load in regulations, technical specifications, and contract documents, and the bots really come in handy with routine tasks,” Jeff says. He’s now working with Pape-Dawson’s Geospatial Information Services to upload existing terrain and infrastructure information to the data that AI consults when rendering its recommendations.

In one recent MUD use, a contractor left the job site without reason or informing the district. The team fed AI the contract and timeline and asked it to evaluate legal liability. “The AI came up with pretty much everything our legal team found, and even some other ideas, too,” Jason says.

Predictive modeling—foretelling the probable future for a MUD—currently has the team excited, says Jason. “A merely good engineer can input flood gage data into a GIS to report on current flood conditions,” he states. “A great engineer, though, can use AI to anticipate what’s coming. Instead of telling residents what’s happening now, we can tell them what’s going to happen in two or three days and how to prepare for that.” Similarly, MUDs can keep an inventory of components for water plants, say, and predict when they will need replacement, as well as associated repair costs. Feeding costs into larger models that account for capital improvements, tax rates, and other inputs can help with budgeting, as well as preventive maintenance.

Vanessa McMahan, Vice President, Innovation is leading the charge on Pape-Dawson’s AI integration. Yet as she moves the firm forward, she maintains a careful focus on old-school engineering skills. “Ours is an apprenticeship industry,” she says. “It’s really important that we use AI in ways that don’t erode our engineering abilities.” Administrative tasks and first-round quality control are areas where AI can help without risking engineering atrophy. With the goal of increasing productivity 20% by 2026, Vanessa’s Innovation Center coordinates across Pape-Dawson to adopt and pioneer the use of new technologies.

When it comes to AI, Jason is enthusiastic: “As we look at new uses, our imagination is the limit.” That’s true for MUDs and for the everyday, hands-on work Pape-Dawson staff do to keep communities growing and thriving.