By Bob LoGalbo, Leidos
Regulated business environments inherently have less room to grow. As a result, exposing savings opportunities is often a route to increased profitability. If institutionalized in a utility, data science can clearly define those savings opportunities for executives. Without data science, utility executives are relegated to less informed and insufficiently impactful decisions. The growing awareness around data science combined with the growing availability of data science prowess offers utilities the opportunity for better decision making and financial growth.
Moving Up the Data Science Value Chain
To some degree, all utilities already incorporate at least the basics of data science. If a chart is generated from Microsoft Excel, strictly speaking, a first approach of data science has been applied. Changing the view of a spreadsheet pivot table is essentially what data scientists do when examining data sets. Excel-based analysis provides insight and informs decision making. This level of analysis is already familiar to any business leader who has scrutinized a chart on a presentation slide. The term data science, therefore, should not be intimidating to the point that it scares people into avoiding what data science has to offer. Given that Excel-based (or similarly generated) data science is already relied upon by utility executives, the latest high processing data science using the latest algorithms should provide only more perspectives, insight and value than what Excel could ever offer.
If only standard hardware and software are relied upon for data science, there is a physical limit to the amount of insight that can be generated. A laptop and Excel have only so much available processing power and memory to wrest out insight. Executives are then relegated to a limited view generated from a constrained architecture. Using cloud computing power and memory are orders of magnitude more cost effective than daisy chaining desktops, and, by running mature data science algorithms (i.e. the definition of Big Data), previously unexplored insight can be flushed out. Some utilities have realized this and are fully committing to investing in big data programs. Others remain on lower rungs of the data science ladder. A utility that simply considers, explores or dabbles with big data risks lagging behind. A utility may lose either a competitive market edge or public support if it doesn’t realize the same efficiencies as the utilities that have matured their big data programs.
Outside, Looking In
When compared to other industries, many utilities risk being classified as technological laggards. This classification damages reputations and can potentially dissuade young talent from considering employment within the utility industry.
More than 400 universities in the U.S. now offer degrees in data science, and many also offer dual degree programs with data science and other degrees such as electrical engineering, business and even anthropology. Dual degree options are becoming not only a more common offering at universities but also a more common choice for incoming freshman. This trend indicates that academia sees data science as vital for not only engineering but for all disciplines. The most capable entry-level candidates know this, which therefore makes dual data science degree graduates a perfect fit for an organization like a utility, that is flush with data. This is true only if the utility is “all in” with respect to data science. If not committed to the benefits of data science, the utility is sure to be passed over by these candidates.
Hazards of Downplaying Data Science
Given the potential strength of data science to significantly accelerate any savings or revenue enhancement, there is a clear risk that not going “all in” might leave the organization in a compromising position. Not advancing a data science program to higher processing power leaves money on the table. More mature data science algorithms wielded by internal or external data science experts can provide significant value to a utility. Downplaying data science can reduce the value of a utility’s data and even lead to poor decision making.
System complexity, regulatory policy and competitive pressure are increasing rapidly, and lower-level data science architecture can be easily overwhelmed by the processing required to accommodate these intricacies. Big Data can align the varied factors and determine the optimal solution to maximize savings. In some cases, a reliance on limited data science architecture (such as laptops and Excel) to unknot these complicated variables might not provide only limited insight, but might actually provide poor options from which to choose, steering executives into making damaging decisions.
In addition, with an aging subject matter expert (SME) population in the utility industry, competing to hire the “best and brightest” is no longer the sole building block to corporate growth. Big Data through machine learning has the potential to capture and execute automated SME expertise. The ability to build infrastructure and code solutions, however, requires adept and adroit “white hat hackers” with a data science background. A commitment to a Big Data program will either attract internal data science talent or allow outside experts to truly provide transformational data science support.
Going the Full Distance to Growth
Utilities have access to terabytes of data that, upon analysis by data experts, can provide not only value from overlooked insight of current analytics, but can be used to generate predictors. The predictors can take a current state and project a future state with a potentially high degree of accuracy and precision. It is those predictors that provide a magnifier of corporate revenue. Even in a regulated environment, the right data providing the right insight will lead to the right decisions. Utilities willing to internally build an advanced data science program or to bolt on outside data science expertise will be strongly positioned for growth.
Bob LoGalbo is chief data scientist at Leidos.
