AI for Predicting Demand in Water Distribution Networks

AI for Predicting Demand in Water Distribution Networks

Ooof! What a weekend its been. 

I wrote 5 articles on AI in Water Tech.

I wrote two personal introspective blog posts - including one focused on accessibility and one ticket at a time. I wrote some urgent notes for work - not something I want to do on a weekend but it had to be done. Next week it will be better. I also didn't get to wrap up properly on Friday and that fed into my weekend - need a clear to do list and tabs closed by the end of the week or else I can't stop. I need to use my standup update workflow tomorrow and I need to be in on time.

So let me just wrap this water tech thing up quickly please - I wanted to just finish one more reflection piece.

AI in Water Distribution Networks

I am so interested in this.

I am interested in AI and how it can be used to monitor and optimise water networks in real time.

It's just to me that there is a lot going on here and I am trying to unpick it and wrap it all up. AI can analyse data from two types of places in real time that tell it a lot about flow:
  • Water meters
    • Home meters 
    • Commercial meters
    • I am not sure if these exist but maybe like they have meters somewhere along the network?
      • Obviously these would need to be smart meters
  • Data from sensors
    • Sensor data is so key to AI in utilities
    • I don't even know where to start
    • I am not sure of all the kinds of places that water networks can have sensors in but this is something to explore 

AI can help...

  • To predict fluctuations in water demand
  • It can spot patterns in consumption
    • So if the patterns are off we know there is a leak
    • Again this is something that my team works on - WOW!!!
  • I am curious about predicting fluctuations in water demand. Hold on

How AI can Predict Water Demand

Data Collection

  • By using data from historical usage, we can use the past to predict the demand for the future
  • By using weather data, we can use temperature, precipitation and humidity to predict things
  • DEMOGRAPHIC INFORMATION - OMG WOW THIS ONE IS SO SO FASCINATING AND SO INTERESTING WOW. 
    • It merits a whole blog post in and of itself 
    • We can look at:
      • POPULATION GROWTH 
      • URBANIZATION RATES
      • AND CHANGES IN HOUSEHOLD SIZE
        • TO PREDICT UPCOMING WATER DEMAND
        • SORRY I AM JUST SO SO SO EXCITED - WOW
  • Economic indicators - 
    • economic activities
    • industrial growth
      • income levels???! 
        • If someone buys a jacuzzi that's more water
        • Also missing from here are major events eg sporting events as with energy
  • Behavioural data
    • If we have smart meter data this can reveal patterns (And data from IoT devices in the home - internet of things - more on this another time).
All of this can be fed into ML models.

This is what we do in our team although I am a software dev so I link models to users if that makes sense.

Monitoring Data from Real Time Sources

  • We can monitor consumption patterns in real time
  • We can detect leaks and inefficiencies in real time
    • We can intervene quickly
  • We can respond quickly.
Predictive analytics allow us to plan for scenarios.

This:
  • Increases accuracy
  • Improves costs
  • Allows us to STORE, MANAGE AND DISTRIBUTE WATER BETTER, AND IMPROVE WATER INFRASTRUCTURE.
  • IS MORE SUSTAINABLE!!! !!! !!! !!! !!! !!!
  • Increases customer satisfaction!!!

Next time

I have more articles to go and perhaps should move on but
  • Still want to reflect on AI in preventing water loss
  • Might want to reflect on the above chapter - everything from `HOW AI CAN PREDICT WATER DEMAND` DOWN PLEASE THANKS.
IMAGES OF WATER AND CODE


I can take the discomfort. I can handle the discomfort. The discomfort that comes with not knowing. What the next article will be on. 

Where the direction of the next research will take me in. 

I just want to know more.

Comments

Popular posts from this blog

Hello World

Where To Hide A Star

In the Water, I find Fire