“IT and OT security are crucial, as water infrastructure must be protected for the public good”

Can you tell us briefly about your career path and your current role at Maynilad?

I have been with Maynilad Water Services, Inc. since 2011, and I am currently Senior Vice President - Chief Information Officer. Previously, I was with a multinational consulting company for over 13 years, where I worked in my last role as Regional Head for Asia-Pacific’s consulting practice. I have an undergraduate degree in B.S. Electronics & Communications Engg. (DLSU-Manila), and a Ph.D. in Electronics Engineering from the Univ. Politecnica de Catalunya (Barcelona, Spain) where I also taught and was the Vice-Director for the Technical Engineering College, conducting R&D as well in collaboration with industry and the EU. I also had a stint in a manufacturing company, in charge of the automation and instrumentation of their production lines, which gave me an understanding of OT, the reason why I now also handle the OT for Maynilad’s plants since 2014.

For the last 10 years, Maynilad has pursued a strategy that involves digitalising the operational technology at all plants and connecting it with the information technology. Could you explain why digitalisation plays such a crucial role for Maynilad and what the benefits are?

Firstly, today’s water treatment and sewage treatment plants are very complex and have a vast number of equipment which needs to be controlled. As such, it is impossible to operate them effectively without the use of Automation (OT), which would only be true for very small, conventional types of facilities.

Furthermore, integrating OT with IT offers numerous benefits and has proven to be advantageous for companies worldwide, extending beyond just the utilities sector, for example, for manufacturing companies, oil and gas, etc. The first and most immediate benefit is the visibility of data to management. OT has the data, but it is in a “raw” form. However, by integrating it with IT, we can use IT tools to turn that into information: dashboards and reports that management can understand and use to be informed on how the business is going. The second biggest benefit is having a single source of truth for technical data. In Maynilad’s case, we started building our IoT (Internet of Things) platform in 2012, with the intention of it acting as a repository of all technical data. Given that over 12 years have passed, this is a rich source of information, which we are now leveraging to do data analytics.

Maynilad has installed its own weather station. Could you tell us more about this initiative and its advantages?

It is also related to the data analytics aspects. When we initiated deeper analytics and predictive modelling, we noticed that most of the problems we had at hand required good weather data. We then contracted a weather company to come up with specific predictions for us; on rainfall, temperature and wind direction, they used available data sources, but in order to obtain more accurate predictions, we decided to install our own Automatic Weather Stations in sites that are critical to us: in our catchment areas in Ipo, Angat and near Laguna Lake. These send to us hourly weather readings via satellite which would not only inform us on the current weather situation, but also feed it in the weather prediction models of our provider, making them more accurate.

How does integrating AI technology into Maynilad's operations contribute to its Non-Revenue Water (NRW) Management Program?

Currently, 80 per cent to 90 per cent of Maynilad’s NRW are due to leaks and up to 90 per cent of these leaks do not surface, which means these need additional efforts to locate. Leakage detection is therefore a critical element in Maynilad’s NRW management program.

Integrating artificial intelligence into Maynilad's operations enhances our NRW program by augmenting our leak detection capabilities

AI technology helps Maynilad to find more leaks at a faster rate. Integrating AI technology into Maynilad's operations enhances our NRW program by augmenting and complementing our existing technologies and leak detection capabilities. By using machine learning algorithms, AI identifies segments of the pipe network where leaks are likely to occur based on available information such as pipe materials, age, history of repairs and NRW data. AI helps optimize the allocation of resources, such as manpower and equipment, by localizing areas of the pipe network with a high likelihood of pipe failure, ensuring that efforts are focused on the most critical areas.

What were the key results of Maynilad's pilot run of the Infrawise AI technology, and how did it impact the company’s leak detection efficiency?

The first run of Infrawise involved 1,700 kilometres of pipelines from 410 district-metered areas with high NRW. Infrawise identified 756 kilometres with a high likelihood of failure. The leak detection teams are then deployed to these segments to inspect and pinpoint the exact location of leaks. Maynilad was able to find 1,525 leaks from these segments resulting in an average output of two leaks per kilometre of inspection.

The rate of finding leaks in the segments recommended by Infrawise is higher than the rate when the entire pipe networks are investigated, commonly termed as blanket leak detection, which is only 1.59 leak per kilometre.

Infrawise is an effective complement to the standard approach of leak localization, termed step-testing. The latter involves the deployment of the team on the ground to investigate and localize areas of the pipe network for focus leak detection.

Can you explain how Asterra's satellite-based technology detects underground pipe leaks, and what advantages it offers over traditional leak detection methods?

Asterra applies algorithmic analysis to track the spectral “signature” of potable water underground captured in a satellite image. The leakage information that the AI algorithms pick up is then captured in a Geographic Information System report that specifies street locations.

Asterra's satellite-based approach allows for the inspection of vast areas in a single scan, covering entire cities or regions. This is especially beneficial for utilities managing extensive water distribution networks. Unlike traditional methods, it avoids the need for physical access to the pipes, reducing disruption and the risk of infrastructure damage. The technology can rapidly assess large areas without extensive groundwork, speeding up the leak detection process.

What are the main challenges Maynilad faces when implementing AI and satellite technologies in its water operations, and how are they addressing these challenges?

The implementation of advanced technologies involves significant upfront costs, particularly for software. These initial expenses can be substantial, but they are necessary for establishing a strong technological infrastructure. Ongoing costs include system maintenance, data storage, and the potential need for upgrading technology over time. The integration of AI and satellite systems, while enhancing capabilities, also increases the potential for cybersecurity threats. Ensuring compliance with data protection regulations and standards is essential, particularly when handling sensitive customer data.

Other critical factors can affect the successful implementation of new technologies, such as the quality of the data being collected and analyzed. Poor data quality can lead to inaccurate insights and ineffective decision-making. Additionally, a sudden surge in areas requiring attention from the leak detection team necessitates careful prioritization of deployment. This may involve adding more teams and equipment to handle the increased workload efficiently. It's crucial to assess the readiness of our operations to manage a higher volume of leak repairs, ensuring that the infrastructure and personnel are adequately prepared for the increased demand. In 2024, the average number of underground leaks detected per month increased by 70%, rising from 1,129 in 2023 to 1,929.

What future advancements in digital technologies is Maynilad exploring to further improve its water operations and efficiency?

One of the focus areas of Maynilad is in customer service and as such, we are investing in new platforms to help in this. Currently, one of our challenges is being able to reply to all the queries we get through social media, especially when we have maintenance or leaks that may affect our service. During these episodes, we have such a high volume of queries that our operators cannot reply. An AI-based chatbot that handles such volume of queries efficiently is already being used. We are focusing on data analytics in general, because thanks to the use of such tools, we can drive more efficiencies in our operations, which translates to lower costs.

We will also continue to automate all our plants and facilities, as well as hooking them up to our Central Control Room and putting all equipment and sensor information into our IoT platform.

Lastly, IT and OT security are crucial, as water, being a vital part of critical infrastructure, must be rigorously protected for the public good.