Water drop

Creating Value through Applied Data Science
P Prevos
Publication Date (Web): 31 January 2017
DOI: https://doi.org/10.21139/wej.2017.008

The words “Big Data” have become synonymous with promises of boundless benefits. Big Data algorithms are attributed almost mystical capabilities to improve the experience of customers or optimise treatment processes and profoundly change urban water management overall.

There are some famous examples of successful companies such as Facebook, Amazon and Google, where Big Data forms part of the fabric of the enterprise. But for most organisations, including water utilities, success in this area has been limited.

The Big Data moniker is burdened with undelivered promises. As Big Data surfs the technology hype curve, the discipline of Data Science emerges as a practical way to extract more value from data. Data Science is a multidisciplinary field that combines mathematics, computing and subject matter expertise to develop actionable insights. Value from data comes from sound, useful and aesthetic information. Data products firstly need to be reliable and useful to add value and aesthetics ensure that information is communicated in a comprehensible way.

Data analytics comes naturally to the engineering and science-focused organisations that water utilities are. Given our reliance on data and technology, the benefits promised by Big Data should be within reach for water utilities.

The Data Science Continuum (Figure 2) illustrates the value chain for business analytics. Before any value can be created, we first need to assure data quality because the law of Rubbish-in-Rubbish-Out is immutable. The Data Science Continuum is strictly hierarchical. Organisations cannot evolve to a level without first mastering the previous ones.
Figure 2: Data Science Continuum
 Figure 2: Data Science Continuum (Mongeau, 2014).

The majority of data management projects are located on the second level of the continuum. Business reports provide descriptive statistics that summarise existing data. Data visualisation techniques enhance the information to help the reader better understand the presented information and make better decisions. The craft of visualising data has developed in the past years combining perception psychology and creative input from artists to improve the way data is communicated.

Most engineering and scientific analyses in water utilities relates to diagnostics. At this level of value, new information is created from existing data. Water utilities mostly operate at this level through water and sewer system modelling.


Predictive analysis or machine learning is the core of Data Science. New predictive analysis techniques provide opportunities to understand our customers and equipment better.

The final two levels of the data science value chain enable the Intelligent Water Network. In prescriptive analysis, the outcomes of diagnostics and predictive analysis are used to make operational decisions. Semantic analysis relates to using analysis to understand the messy data of the social world better.

Coliban Water has enacted a strategy that allows us to reap the benefits of the emerging data science technologies to improve the customer experience and bottom line of the organisation. This essay provides a primer on how to leverage the benefits of new developments in data science to improve water utilities. Read the full article for examples and practical advice on implementing data science projects by leveraging existing technologies and competencies.


Click here to read the full paper

Not a member? You can purchase the paper in our Online Document Library.