The Catalan Water Partnership (CWP) has started the second season of its knowledge pills. These are short videos on the trends, technologies and challenges that are changing the water sector.
The first episode looks at the impact of artificial intelligence. It brings together three views: CICLAGUA, AVENTEC and Earthpulse. Fran Martín Rivas, our Product Manager, made three points. Here is a short summary of each one.
1. AI makes water data easier to use
Water data is complex. Typically, a full water network runs with many types of sensors and many data sources. While in the past, using this data took many technical profiles and many steps — especially as someone expert had to access the data, process it and then extract the value — now this is different. AI has, in fact, lowered access barrier. Today, with AI models curated and designed to do this job, it is much easier to find patterns and predict risks.
At Earthpulse, our central initiative is generative AI. Any user of our platform, SPAI, can ask SPAI chat a simple question, for example: where is the risk of water leaks highest? Where is water quality the worst?
This system is designed with user and usability in mind from the first step. The user gets the value of the data much sooner. It also has expanded use cases and how the data can be used.
2. The hard part is moving AI into operations
It is generally considered easy to find a good AI model in a controlled, scientific setting, but we know it is much harder to run that model in real operations.
Working with clients, it is clear that real operations need processing in near-real-time. Operations also cover different areas, different geographies and different use cases. Building a good model and a good service at this level is the real challenge, but once the infrastructure and the knowledge are in place, the benefits are clear.
The biggest benefit is scale. You can monitor very large regions, even the whole world and in a matter of minutes you can have data and analytics and make decisions faster.
3. Earth observation and AI scale where sensors and field visits cannot
Traditionally, water companies and public administrations manage large networks over wide territories by relying on sensors and field inspections. Both work well, but neither one scales easily.
The difference is brought by how quickly Earth Observation data combined with AI changes this. In real terms, it lets operators monitor large areas without difficulty in near real time. This way, operators need fewer field inspections, or more targeted field inspections to double check accuracy of findings. The other benefits being that they can see all the information in one control panel.
Watch the full video
Fran speaks at 2:00, 4:37 and 6:04. You can watch the full episode here:
Thank you to the Catalan Water Partnership for the invitation.
Thank you also to CICLAGUA and AVENTEC for a useful conversation.
