"The main barrier to water digitalisation is not technological, but organisational & generational"

Many utilities already have advanced digital systems, yet we still see decisions being made with logic very similar to that of the pre-digital era. Where do you think the main bottleneck lies that prevents all this information from translating into a different way of operating services?

The digital transformation of water is not solved with advanced algorithms, but with the real integration of system data

For me, the bottleneck lies in how the generational change is taking place within utilities. There is a generation with tremendous knowledge of networks, assets, and plants, from whom I myself have learned a great deal. As those people retire, a new generation comes in expecting the digital solution to directly provide operating setpoints and improvements to apply.

But in most cases, the prior work hasn’t been done. Two or three years haven’t been devoted to integrating all the utility’s data. What we see instead is a strong focus on SCADA, geographic information systems, or hydraulic models, both for drinking water and wastewater, and then very specific, sometimes quite manual, cross-referencing of sensor data.

That allows for some operational suggestions, but it doesn’t represent a model change. What we see are pilots, experiments, and very specific optimisations in different areas. This happens even among early adopters, and in many other utilities, they don’t even get that far.

It’s clear that these digital solutions have become increasingly powerful and complex, but perhaps what’s missing is sound judgment in how they’re implemented. Based on your experience, if you were starting to digitise today, what essential criteria should be carefully considered to decide what to digitise, how far to go, and what not to touch for now?

First and foremost, you have to work on the data model. It’s essential to combine the different traditional sources of information, ERP, SCADA, and geographic information systems, with new data coming from IoT and distributed sensors, and to do it properly: not just to handle large volumes of information, but to align it with the business model and the full water cycle.

That integration also needs to extend to the commercial and billing system, which is always highly complex, as well as to operational systems. This requires people with a great deal of know-how about the water business, and professionals with that profile are not very common.

Additionally, it’s very common to find utilities where teams work in partially isolated ways. Even though in some cases there are combinations across areas, the water cycle is not truly reflected in the company’s data model, which makes growth and the generation of efficient management dynamics extremely difficult.

You are a global technology provider operating in a very fragmented sector, with regulations that vary widely between countries and even within the same country. To what extent are your solutions applicable “out of the box” when they reach a utility, and to what extent is it necessary to tailor them to each operational case?

One of the things we’ve discovered working in different countries is that the world is more similar than we often think. It’s true that regulations change and are usually the main factor introducing differences, because they shape needs: reporting levels, transparency requirements, or the indicators that must be met.

For example, in wastewater treatment, KPIs differ by country; in some contexts, regulatory demands are very high, and in others, they’ve intensified recently. In that sense, regulation clearly marks differences.

However, when you go down to the deep operational level, processes are practically the same everywhere. Assets may differ — desalination plants, wells, or drinking water treatment plants — but in the end, different technologies are used to solve very similar problems. That’s where the engineer adds value, adapting treatment to each specific reality.

If you look at the whole picture, the common operational ground is quite broad. Differences exist, but the operational base on which digitalisation is built is surprisingly common.

We’ve talked about your global presence and the challenges you face in this new phase, even greater than when you operated solely as Idrica. I’d like you to tell us about that journey — from the birth of Idrica out of Global Omnium’s experience to your current position at Xylem Vue. When did this integration arise, and what has really changed for you?

For us, the change has been very significant. First of all, doors have opened that were previously very difficult to access, and that has to do with being part of a large industrial company like Xylem. From there, we started working on projects that were difficult to tackle when we were just Idrica.

Moreover, this didn’t arise from a one-off transaction. Even before the integration, we were already working together on projects where there was clear synergy. That relationship grew until Xylem became the majority shareholder. Today, we are part of the group, and the global digital solution offered to the market under this approach is Xylem Vue.

With experience rooted in a utility and now integrated into an industrial group like Xylem, your relationship with service operators is very close. In that context, how is that relationship structured around data governance and decision-making? Where does your role begin and end?

Today, I would say that, even though many organisations in the sector haven’t yet completed their digital transformation, practically all of them are clear that they must embark on a data governance journey. In every operator we work with, there is already someone responsible for data governance, and when that role doesn’t exist, it’s created as soon as a digital transformation project begins.

Data governance is the exclusive property of the service operator. We do not govern the data. What we do is help, based on the knowledge we’ve accumulated, to build the data governance structure within the organisation. Our solutions accelerate that process, but always with a very clear objective: that the operator becomes autonomous.

We support organisations so they can adopt the technology and knowledge needed and, within one or two years, can make decisions on their own using the platform. It’s a tough process at the beginning and generates internal friction, but once it’s underway and results start to appear, operations managers are usually very proud of what they’ve built.

And it’s important to clarify: we don’t operate infrastructure, and in many cases we don’t directly optimise either. Xylem sells technology; it doesn’t operate. We work with engineering firms and consultancies that carry out implementation and optimisation using our technology. Our role is to facilitate, accelerate, and provide autonomy.

In the digital adoption process, one of the major shifts is moving from reactive operations — the historical norm — to more proactive operations. In a context marked by increasingly frequent extreme events, higher regulatory demands, and growing transparency requirements, are we moving from anticipation as an operational improvement to anticipation as a structural necessity in water management?

From an operational standpoint, you can distinguish three models. The first is corrective or reactive, where you act once the problem has already occurred. The second is preventive, where actions are taken to prevent something from happening. And the third is predictive, where you try to anticipate what will happen and act beforehand.

In today’s context, with increasingly aggressive weather phenomena, anticipation stops being a desirable improvement and becomes a necessity. Anticipating events and sending information to civil protection and public authorities is something many organisations aspire to today, but with climate evolution, it will become mandatory.

This has been our perspective for years. That’s why within Xylem, there is a specific unit dedicated to managing extreme events, with a differentiated approach, because we understand this area requires its own capabilities. We believe predictive management of floods and droughts will be a key component of water operations in the coming years.

So far, we’ve talked a lot about utilities, which are clearly a fundamental segment for you. But your solutions can also be applied to other areas, such as irrigation, industry, or integrated basin management. How do these other users fit into your day-to-day work, and how do these solutions coexist?

Xylem is an industrial company, which naturally places us in many areas beyond utilities. In industry, for example, projects often begin by addressing water management within a facility, usually treatment, and from there expands to other processes, because the platform allows information from different systems to be related.

The challenge of integrating Idrica into Xylem has been to unify different working dynamics without losing technological agility

That approach isn’t new to us. While we were developing solutions from Idrica for utilities, we were simultaneously working on solutions for industry. The technological base is the same: a data model capable of integrating information from multiple sources and turning it into operational support. What changes is the context and the specific process to be optimised.

Irrigation is a particularly relevant area due to its share of water consumption. We have specific solutions aimed at irrigation communities, and it’s a sector where we’re clearly growing. There, digitalisation has a direct impact on both resource efficiency and system sustainability.

And basin management follows a similar logic. The idea is to integrate information, anticipate scenarios, and support decision-making on a broader scale. Ultimately, all these solutions coexist because they share the same technological foundation and philosophy: using data and digitalisation to improve water management in different contexts, but with very similar underlying challenges.

We’re experiencing enormous momentum around artificial intelligence, and no sector is immune to it. I wanted to ask you two things. First, how are you integrating AI into your solutions, both through your own developments and external models? And second, to what extent do you think this technology can replace people in the daily operation of water services?

Our commitment to artificial intelligence isn’t recent. When we were still Idrica, we were already clearly incorporating machine learning into our solutions, and we’ve been working with these kinds of algorithms for years. It’s not something we’ve added just because it’s fashionable.

With generative AI, what we’ve done is build an architecture so that AI always works within a well-defined context. We have a standard data model for end-to-end water management, which we’ve connected to an MCP server (Model Context Protocol) that allows AI agents to query structured information, using both proprietary and external models. In addition, each utility contributes its own knowledge context, so responses are aligned with its operational reality.

In this whole process of data acquisition and processing, automation, and decision-making, there’s one variable that must permeate the entire strategy: cybersecurity. Do you think the water sector is addressing this challenge maturely across all its critical assets? And in that sense, does technological progress inevitably imply greater risks?

Risks certainly exist. But I wouldn’t say the sector is approaching cybersecurity immaturely; rather, I’d say it’s taking it very seriously. Today, it’s hard to find a water company that isn’t treating this issue with great seriousness. Moreover, governments and regulators in many different countries are driving clear regulations and requirements in this area.

That said, we do see two very different approaches within organisations. On the one hand, there are models where cybersecurity acts almost like a censor, focused exclusively on blocking risks, without considering how that affects the business and operations. Taken to the extreme, that approach can end up slowing down or even paralysing digital transformation processes.

On the other hand, there’s an approach where cybersecurity is aligned with the business, understanding the need to protect systems but also to enable operations and innovation. In this model, controls and limits are set when necessary, but always with a strategic vision. That difference in attitude completely changes an organisation.

In my view, cybersecurity cannot be just a technical function. It has to become a strategic function, capable of understanding technology as well as the business, operations, and the company’s overall objectives. When managed that way, it stops being an obstacle and becomes an enabler of digital transformation.

We’re coming off several years of very strong investment in digitalisation, especially after the COVID-19 pandemic. Do you see this trend continuing, accelerating, or is there a risk it could slow down? And in that context, how do you assess the impact of the PERTE in Spain? Do you think this model can be exported to other countries?

Investment in digitalisation is clearly happening in the sector beyond the pandemic, although it’s true that the pandemic was a turning point. Even so, globally, utilities have been slower than one might expect in making that investment. That said, the trend is clear: the water sector cannot afford not to make the digital leap.

We’re also seeing something that concerns us: some companies that were our direct competitors are stepping back and moving away from digital water to focus on other technological areas. It doesn’t usually make big headlines, but there is a gradual divestment in the sector, and that’s a warning sign. Water needs specialised companies and sustained investment to move forward.

As for the PERTE, despite frustrations and significant bureaucratic burdens, I believe it’s been a very positive model and has clearly stimulated investment in Spain. We’ve experienced it firsthand and are involved in several projects. We also believe it’s a replicable model, and we’re already seeing some countries draw inspiration from this experience to drive their own water sector digitalisation.

Finally, I’d like to ask for a forward-looking message. Where do you think the water industry is headed over the next five or ten years, and what role will Xylem Vue play in that process?

The water industry is moving toward a much more intelligent and connected model, where digitalisation will be key not only to optimising operations but to making proactive decisions. The goal is to manage water more efficiently and sustainably, and to anticipate problems before they occur. What we’ve experienced in recent years with extreme events is just a preview of what’s coming, and the ability to anticipate will be decisive.

In that context, artificial intelligence will play a very important role. We’ll see utilities and other sector players, as well, incorporate AI not only to manage water, but to better understand its impact in areas like agriculture or industry. AI will enable faster decisions, based on processed data and clearer business logic.

As for Xylem Vue, our role is to be an enabler of that change. We want to be the digital platform that allows utilities and other stakeholders to manage the water cycle more efficiently, anticipate what’s going to happen, and make better decisions. Our focus is on accompanying the sector through this transformation, providing technology and knowledge to address the challenges ahead.