Potential application areas for further work as more smart data becomes routinely
available include:
• Data mining: understanding how businesses and households use water and
whether and where unique patterns in this use exist is essential for proactive
management (including weekday vs. weekend analysis).
• Applying customer segmentation based on consumption data for customer
loading and variable water pricing in a similar manner to energy (some industrial
customers already have tariffs based on time of day usage).
• For leak detection activities based on detecting pattern changes (deviation
from cluster centroids/distributions); data-driven models of demand could also
help identify atypical customers or unusual changes in consumption.
• Filling missing data for audits/regulatory purposes using cluster centroids/typical
usage perhaps allowing volumetric usage and flow profiles to be estimated for
unmetered customers.
• Allowing a WSP to forecast the demand of a client or the overall DMA and
hence improved network operations.
• More accurate demand profiles for hydraulic modelling.
9 Discussion of Technology Adoption
and Recommendations
Water is considered a ‘right’ in many parts of the developed world which
have grown accustomed to clean drinking water and sanitary facility provision.
Many of the world’s leading water companies have been around for decades,
even centuries. They enjoy high prestige, low staff turnover and healthy margins.
Few see threats from competitors poaching metered customers who are tethered
to miles of buried pipe infrastructure. Regulation can also be a barrier to innovation.
As natural monopolies, the incumbent water utilities generally feel safe and insulated
from competition. Digitisation was once seen as a luxury. However, innovators are
disrupting the old business models, and there is little place anymore for complacency
with threats such as decentralised and distributed technology arising. Disruptive
innovation need not be a zero-sum game in which only one side emerges victorious.
There is no reason why water utilities cannot learn from insurgents, engage with
new thinking and embrace innovation to update their business models to deliver new
solutions that benefit all.
A further issue is that the water sector is not generally perceived as a ‘cool’
industry, partly due to it not being at the forefront of the technology adoption
curve. In contrast, new digital technologies are a hot topic particularly as machine
learning and AI begin to proliferate into industrial application. This fact means
that a career in the water industry generally isn’t a top priority for data scientist
professionals and attracting the right type of talent can be difficult. Startups in the
water sector can rarely compete financially with Google for hiring coders, graphic
designers and tech engineers. Nor can potential return on investment compete
Data Science Trends and Opportunities for Smart Water Utilities
21
available include:
• Data mining: understanding how businesses and households use water and
whether and where unique patterns in this use exist is essential for proactive
management (including weekday vs. weekend analysis).
• Applying customer segmentation based on consumption data for customer
loading and variable water pricing in a similar manner to energy (some industrial
customers already have tariffs based on time of day usage).
• For leak detection activities based on detecting pattern changes (deviation
from cluster centroids/distributions); data-driven models of demand could also
help identify atypical customers or unusual changes in consumption.
• Filling missing data for audits/regulatory purposes using cluster centroids/typical
usage perhaps allowing volumetric usage and flow profiles to be estimated for
unmetered customers.
• Allowing a WSP to forecast the demand of a client or the overall DMA and
hence improved network operations.
• More accurate demand profiles for hydraulic modelling.
9 Discussion of Technology Adoption
and Recommendations
Water is considered a ‘right’ in many parts of the developed world which
have grown accustomed to clean drinking water and sanitary facility provision.
Many of the world’s leading water companies have been around for decades,
even centuries. They enjoy high prestige, low staff turnover and healthy margins.
Few see threats from competitors poaching metered customers who are tethered
to miles of buried pipe infrastructure. Regulation can also be a barrier to innovation.
As natural monopolies, the incumbent water utilities generally feel safe and insulated
from competition. Digitisation was once seen as a luxury. However, innovators are
disrupting the old business models, and there is little place anymore for complacency
with threats such as decentralised and distributed technology arising. Disruptive
innovation need not be a zero-sum game in which only one side emerges victorious.
There is no reason why water utilities cannot learn from insurgents, engage with
new thinking and embrace innovation to update their business models to deliver new
solutions that benefit all.
A further issue is that the water sector is not generally perceived as a ‘cool’
industry, partly due to it not being at the forefront of the technology adoption
curve. In contrast, new digital technologies are a hot topic particularly as machine
learning and AI begin to proliferate into industrial application. This fact means
that a career in the water industry generally isn’t a top priority for data scientist
professionals and attracting the right type of talent can be difficult. Startups in the
water sector can rarely compete financially with Google for hiring coders, graphic
designers and tech engineers. Nor can potential return on investment compete
Data Science Trends and Opportunities for Smart Water Utilities
21