This article was taken from the paper “AI & Water management — What utilities need to know now.” You’ll find the full paper here.
AI often gets imbued with the sense that it’s going to solve everything and work everywhere. And while optimism about AI’s potential in the water sector is important, it’s best weighed with a healthy dose of discretion, and a focus on the unique goals and priorities of individual utilities. If utilities are interested in AI, it’s important to learn more about AI applications and carefully consider the problems you need to solve, so when vendors come to your organizations with big AI promises, you’re ready to evaluate whether it’s a solution that truly aligns with your goals.
The deployment of any AI solution should be preceded by a clear understanding of the problem or opportunity you have and an assessment of the best potential applications of a given AI solution.
In other words, start with identifying your problem first, then work backwards to the appropriate tool, as opposed to starting with a solution in search of a problem.
There are many tools available to the water sector that use machine learning to help utilities solve problems, and we’ve highlighted some of those applications above. As the excitement around ChatGPT continues, more applications will be developed that are relevant to the water utility sector. Here are some considerations you may want to keep in mind.
One of a utility’s greatest assets is its data. A utility’s data can be incredibly diverse and rich but also spotty with gaps and of uneven quality while also spread out across an organization. There may be customer data, usage rates, service interruption data, financial data, weather data, social data and then of course the many data about network behavior and network assets. The problem is, these data are often siloed between business units, especially in larger organizations. Typically, organizations aren’t fully leveraging their data across business units, a common problem across many industries.
There can be significant opportunities from viewing data as an asset and putting it to work as an asset system to generate insights that can be leveraged to improve decision making and customer experience. Getting to the point where the data can be the engine that feeds and drives the AI will likely require significant action on data hygiene and data preparation.
Cultural fit and change management
The best ideas can die on the vine if they’re not properly nested within the dynamics of the organizational culture, and that applies to the adoption of AI tools as well. Any kind of big change, especially technological change, has a change management implication to it that the water sector needs to acknowledge and keep in mind.
Technological change in an organization can fail and become exceedingly costly for a number of reasons, like an organization not clearly identifying and communicating why a new technology is needed and creating a clear roadmap for measuring its success. Poor change management can also exacerbate employee resistance to change, especially with AI-enabled tools that often bring up fears around the threat of automation and job elimination in the workforce.
But while new technology can repel people, it can also attract people. Utilities may want to consider their future in the wake of the “Silver Tsunami,” an aging demographic and a wave of retiring workers, and whether a new generation of talent will want and expect AI at work.
Do water organizations that resist AI technology risk not attracting the new talent that the water sector will unequivocally need? It’s still perhaps up for debate whether younger generations will choose organizations that embrace AI over others that keep it at arms length. The point is, the message a utility conveys to both its existing and future workforce regarding its utilization of emergent technologies is something to consider.
Small utilities may be more agile
While AI may seem more daunting for small utilities, smaller utilities could have the advantage when it comes to adopting AI tools faster than larger organizations. Simply by virtue of being small, there could be less siloed data, and fewer people can mean easier left-to-right visibility into the organization’s workings, which would make change management less of an undertaking.
In other words, while larger utilities may naturally have more data and analytics to feed and drive AI tools, they could be less agile when it comes to actually mobilizing their organization to adopt new technology. A smaller utility, even if they have a data gap, could be in a better position to bridge any those gaps quickly due to less red tape and less people to align towards the desired technology change.
Nevertheless, utilities of all sizes should approach this new era of artificial intelligence with a sharp eye on the problems they wish to solve, along with related considerations, as they explore options for selecting and deploying relevant AI solutions.
Qatium is co-created with experts and thought leaders from the water industry. We create content to help utilities of all sizes to face current & future challenges.
Paul Fleming, President of WaterValue and Water, Climate and Tech Advisor, is a Qatium advisor.