Skip to content

🤖 "Tackling Critical Water Challenges with Artificial Intelligence & Machine Learning"

Solving Water

Photo by Lucas Kapla / Unsplash

Table of Contents

Host: Amanda Holloway
Guest: Brandon O'Daniel | Data Lake & Data Science Manager | Xylem
Category: 🤖 Technology

Podcast’s Essential Bites:

[8:28] “Our data lake today is used for archival and storage of a lot of information that comes off of a lot of our devices that are out in the field. What we do is we take that data and we mine it, in order to learn things about those devices, to learn how to make them better, and how to improve them over time. Also, we use the data lake to go through and combine all types of different data sources that come in. So for instance, we may take field data that comes in for devices and combine it with say publicly available data like data from the EPA, or data from the Energy Institute of America, […] American Gas Association, or other types of different publicly available datasets.”

[12:57] “The types of applications you would use on anything that's a very data centric type application. So if you have a wastewater application, that is data centric, this could be a prime candidate. […] Smart metering, AMI type of project would be a prime candidate. So it's any place where you have large amounts of data that come in that need to be processed. Also, you can use these types of things in more operational support, for instance a smart chatbot.”

[14:10] “[For utilities] a big area […] [is] predictive maintenance. So being able to go through and figure out all the different data centric factors in a particular device or a particular piece of equipment, that you want to predict, [for example] when the failure may occur sometime in the future […] [to then] schedule […] maintenance cycles.”

[18:00] “This past year, COVID has affected everybody in the world. COVID has actually had an interesting effect on certain types of AI and machine learning, because distribution of data changes. So, for instance, it may be that you are tracking certain types of usage […] in certain geographic areas, and you don't have people with the same migratory patterns that occur from year to year, because they're not moving due to COVID. So you have to adjust and monitor those models and adjust in order to make sure that you're still making the accurate predictions, and you're still providing a valuable service.”

[24:16] “We have a product that's in controlled launch right now […] that's actually titled Utility Data Lake, where a customer can go through and be able to access their own AMI data that comes in off of our flex net systems. So and there are plans in the future to expand out different functionalities to help customers in terms of managing their own data and in their own little private data lakes, so to speak.”

[31:34] “I think the most important thing that I've learned in the water business so far, or at least the part that was very interesting to me, is how different water can be geographically across the world. […] You look at water consumption patterns […] and you can see […] different geographic shifts of how people move around relative to water. And to me, that's just been the most fascinating thing of seeing all these different things about how we really are […] deliberately dependent on water.”

Rating: 💧💧

🎙️ Full Episode: Apple | Spotify
🕰️ 34 min | 🗓️ 06/07/2021
✅ Time saved: 32 min