Bots in the cloud, boots on the ground

Can artificial intelligence and human experience join forces in the future of utilities? It already is.

AI-generated image of a wastewater treatment plant worker and a friendly bot buddy walking hand in hand.

I could have prompted an AI to write this story. Maybe I did.

With the right algorithm, requirements, and patterns, any number of artificial intelligence-powered copywriting tools could spit out a story about the future of AI in the world of wastewater treatment.

But what would be lacking in that computer-generated copy are the years of personal experience of people like Bob Meholif, Brian Villers, and Francis Cung, three Northeast Ohio Regional Sewer District employees who know first-hand how AI and chatbot innovations are — and aren’t — working in the very wastewater treatment services we provide every day.

Bob, Brian, and Fran work where humans and machines meet. Bob is our Manager of Process Controls and Automation, a division that optimizes the performance of our treatment and collection systems through the right combination of human teams and state-of-the-art technology. Brian manages Systems Development and Operations, developing app and data solutions that save time, maximize accessibility, and improve integrations between our many departments. And Fran, our Manager of Data Infrastructure and Business Intelligence, oversees a team of web devs and the systems that collect data of all kinds from all across the Sewer District.

They’re not using AI to generate memes, music, or tweets, but we did ask them for their thoughts on artificial intelligence and its applicability to our industry.

“AI is spreading and adapting in the industrial environment,” Bob said. He spoke of a recent pilot project in our Operations & Maintenance departments using intelligence he called “machine learning.”

“It can be both for equipment control in plant operations and predictive analytics in maintenance,” he explained. “While the results were positive, the staff adaptation and costs weren’t a good fit for us at the time. As the space matures, AI add-ons will become common place.”

“AI is spreading and adapting in the industrial environment.” — Bob Meholif, Manager of Process Controls and Automation

Automation has already enabled each of our three wastewater treatment plants’ processes to be operated, monitored, and adjusted by crews of two or three operators from within centralized hubs. Remote sensors throughout our collection system alert managers of flow levels, regulator actions, and other factors that require human intervention.

These integrations, years in the making, are exactly why Bob uses the word “add-ons” rather than “replacements.” Our utility will always rely heavily on tradespeople and human experience to staff the systems that manage and treat wastewater, skill sets that we’re actively recruiting for the foreseeable future. Artificial intelligence can’t wholly replace that.

But can it compliment? Brian thinks so.

“I agree with Bob,” Brian said via email. “AI is still in its infancy of use and adoption here at the Sewer District. But we have been slowly incorporating the use of ChatGPT into our efforts to enhance development productivity.”

A wastewater plant operator walks between aeration tanks at the Southerly Wastewater Treatment Center in Cuyahoga Heights, Ohio.

Much of Brian’s team’s focus is on developing apps and connections that help departments better integrate, collect, analyze, and report shared data. That requires a lot of coding and testing — things chatbots can help fast-track in early design-stages.

“We’ve used ChatGPT to enhance our code’s documentation. We can drop our code into the chat and request the AI to add comments automatically to explain what each step is doing. Adding comments to code is tremendous for documentation of an application and makes things much easier when coming back to troubleshoot or add additional functionality later.”

Brian told me adding comments to code manually was often overlooked and therefore not always implemented. “ChatGPT has helped automate this process, thus improving all our applications’ sustainability.”

“We can drop our code into the [AI chatbot] and request the AI to add comments automatically to explain what each step is doing.” — Brian Villers, Systems Development and Operations Manager

“ChatGPT can code,” said our Manager of Data Infrastructure and Business Intelligence Francis Cung in a recent Clean Water Works Podcast episode. “You can provide a scenario and it can develop a whole block script, which can help us as a developer,” saving time, resources, and redundancy. Our Information Technology crew knows AI comes with risks and evaluates opportunities accordingly.

Internally and operationally, artificial intelligence is getting a closer look. Externally, even as I’m typing this article, we in communications are dabbling in chatbots and AI image generation not to replace human hands and voices, but to explore new ways to connect with customers. To write more clearly, to spur creativity, to compliment skilled staff people.

Utilities will always require human hands. You’ll still talk to human beings who care when you call Customer Service. But with good data and sound strategy, those hands and voices can help artificial intelligence generate real value for our customers.

“We are just beginning to explore the potential of AI in our day-to-day jobs,” Brian said. “I believe it will continue to improve staff’s productivity in the future.”

AI-generated image of a friendly bot alongside a wastewater treatment plant employee. Cute little thing, but this future may only exist in AI-generated images.

Editor’s note: Yes, a human wrote this story. A bot generated the header and footer images.



Northeast Ohio Regional Sewer District

Official Medium channel of the Northeast Ohio Regional Sewer District in Cleveland, OH