Tim Medearis
It’s always amazed me that all North American water utilities essentially do the same things – cleanse, distribute, and collect water – but the methodologies they use vary widely. If you look at almost every other established industry, there’s often a set or standardized way of solving problems or serving customers. When one competitor makes a move, everyone else moves in unison in relatively short order.
Not so in water, and never is this more on display than at big national conferences like AWWA’s ACE, which our team just returned from. Utilities range size and are dealing with issues like aging infrastructure, workforce development, changing climates and demographics, and varying political and economic realities. And we saw an incredibly diverse amount of solutions and methodologies at ACE.
Within all this diversity and creativity, a few common themes did present themselves.
#1: You can’t skip steps
The buzz around Machine Learning and Artificial Intelligence (ML/AI) was palpable in Denver. Utility directors, in particular, expressed eagerness to learn how to implement these technologies. However, session after session highlighted a crucial and sometimes sobering truth: Defining the problem, gathering data, and preparing this data for machine learning constitutes ~70% of the work in creating an effective machine learning application toolkit. This isn’t the glamorous part of the process; in fact, it’s often tedious and can be expensive, but it’s indispensable.
The “Optimizing Utility Operations Using AI & Machine Learning” session highlighted the needs of gathering and preparing data explicitly, as seen in the screenshot below.
Tacoma Water had another great presentation, and they made this same point indirectly. Their data-intensive hydraulic model was necessary to calibrate and correct IoT data that is too often the sole bedrock of AI/ML tools. I’ve worked on hundreds of hydraulic models, and I know that they require a lot of high-quality data. Although it may take time, effort, and money – it’s worth it. Efforts like hydraulic modeling and quality data collection are the stepping stones to responsible digital transformation.
The reality is that digital transformation is a journey, not a sprint that Chat GPT can perform for you in a matter of seconds. Technologies like AI and ML can provide significant benefits, but they require a solid foundation to build upon. This foundation includes accurate data collection, comprehensive hydraulic models, and a deep understanding of the system’s intricacies. Skipping these steps can lead to unreliable results and missed opportunities for optimization.
#2: Find a way to make it a $1,000,000 problem
All ACE attendees were in Denver because they want to create higher quality, more efficient water services for their communities. One way they see to do this is via digital transformation. However, this change can be expensive. And here is an interesting way to think about that: Expensive solutions require the identification of expensive problems.
Two Colorado Front Range utility engineers I met demonstrated this well. Both utilities had hydraulic models, yet only one – let’s call them “Utility A” – was keeping it up-to-date so that it could form the backbone for future digital advancements. Utility A’s hydraulic model was a key piece to sustainable growth for hundred-million-dollar land development efforts. Utility B’s engineer wanted to be more involved in digital transformation and find ways to use their hydraulic model, but they were struggling to find traction internally for how to do so. Autodesk, our value-added resellers, and our consulting partners all talked with Utility B’s engineer at our booth regarding the many options and resources to help them make their case for why this effort would be so worthwhile.
InfoWorks ICM is widely used for large scale, city wide or larger modeling scenarios, but there are lots of other million-dollar problems that go beyond analyzing inflow and infiltration to complete the relatively small goal of pump selection within lift stations. You can apply hydraulic modeling to lots of million-dollar problems.
For example, it can be used to efficiently retrofit and expand a wastewater treatment plant like Freese & Nichols did for Texas utilities, which we discussed in a Water Talk video. develop and calibrate hydraulic models of wastewater treatment plants to perform capacity assessments and aid in designing new treatment processes for Texas utilities, using it to simulate real-time controls and component designs.
A very similar example could be made for water distribution models. Many times these models are solely used as information for master planning and fire flow analysis. However, as we’ve seen with our recent InfoWater Pro update, you can go further and use your hydraulic models for leakage analysis to more accurately pinpoint where the highest ROI might be found when looking at main replacements. By more effectively minimizing clean water leaks, utilities can not only better mitigate chemical and pumping costs, but also reduce the need for distribution plant upsizing. In essence, practically anything dealing with water plant efficiencies = $1,000,000 problem.
#3: It’s important to distinguish between responsible vs. irresponsible AI
The conference also highlighted the surge in AI/ML tools for many applications, but specifically leakage detection. A utility representative I met noted the overwhelming number of these tools that are already available, and yet he complained that these same product representatives could not sufficiently explain their workings. This raises concerns about the responsible use of AI. While AI holds great promise, it’s essential to ensure that these tools are reliable and their data quality is impeccable.
This is certainly the approach we at Autodesk water solutions are taking. I would have loved to have told a number of attendees that our solutions can use Autodesk’s advanced AI/ML tools to simply automate away all of your hydraulic model creation. I’d love to say that we always have a quick fix and can delight engineers with innovative tools. When we can do that, it is the most fun part of my job! But making marketing claims has to be balanced with responsibility, which is why Autodesk has settled on trust principles for AI.
But overall, talking to water professionals about AI and ML during ACE 2025 was enlightening. Autodesk’s Info360 SaaS-based solutions like Info360 Asset and Info360 Insight generate more enthusiastic interest year after year at ACE due to their inherent cloud and digital twin capabilities. Many water professionals want to invest in cutting-edge, digital twin solutions, and many of them see the potential of going beyond working with desktop software, but they also have to balance their enthusiasm with what their organizations can tackle, which might depend on their overall organization’s digital maturity.
But there’s no denying there is an appetite for new tech solutions. I heard a lot of positive feedback from customers who want to see big data solutions from Civil 3D, Revit, InfoWater Pro, InfoWorks ICM and other utility datasets. They want to get as much use out of SCADA and GIS and bring their work order management into the cloud so they can apply AI/ML solutions effectively. This is one reason Autodesk is on a journey to get as much information from our 200+ existing solutions and our water-specific solutions into the cloud as possible.
Until next year…
Navigating the path to digital transformation for water utilities surely comes with its fair share of challenges, but the payoff is huge. By recognizing the importance of laying down solid groundwork, making strong business cases, and using AI responsibly, the water sector can tap into advanced technologies to achieve cleaner and more efficient water management.
What else did we talk about with customers at ACE 2025?
- What’s ahead for the Autodesk Water Infrastructure portfolio in 2025? Watch the recorded webinar.
- We were happy to announce new product options for InfoWorks ICM customers: Meet the new, flexible options for InfoWorks ICM: Sewer, Flood, and Ultimate
- That wasn’t our only big product news. Visitors to our booth were curious about the re-release of some old software favorites like XPSWMM and InfoAsset Manager. Read about Everything new in Autodesk Water Infrastructure’s 2026 software – plus five new subscription offerings
If I missed you last week and you’d like to know how Autodesk can help you with your digital transformation, please reach out!