Data science for utilities: opportunities and challenges

Every utility company is aware that it should do “something” with data science. The energy sector has no time to waste: the sector is changing rapidly and providers need to make their processes more flexible right now. In this effort, data science plays an essential role. Railroad companies and water suppliers are working hard to reduce costs and guarantee quality. Also for them, data science can be extremely helpful.

In some respects, the challenges facing utilities are no different from those in other industries. A good example is predictive maintenance: closely monitoring the infrastructure makes it possible to predict when some component will fail. This information saves both unnecessary replacements and unexpected failures. In the case of maintenance, the assets in the process industry are not very different from a water supply network or a power supply network.

But utility companies also have challenges that do not occur as clearly in other industries. For example, the customers of utility companies form a much larger and diverse group than in other sectors. The clients often have a legal right to be served but do not have an obligation to give an advanced notification of their demand. Trying to get a clearer picture of the customer requirements may even raise privacy issues.

These customer aspects are rather specific for utility companies. On the one hand, utility companies will therefore (more than other industries) have to make an effort to predict demand and to act on those predictions. In this light, data science is an important tool. On the other hand, utility companies can also use data science to improve customer sentiments by monitoring social media.

It is clear that data science provides all kinds of opportunities for utility companies. But where to begin? Start by formulating a policy and then implement it, even if you do not have any experience yet? Or just start somewhere to gain experience? But wouldn’t that be too time consuming? And wouldn’t it be a waste of money to start doing something without a proper strategy?

Suppliers of data science tools are ready to help utility companies with these kinds of challenges. They offer software that produces impressive figures in just a few clicks, making data science seem almost trivial. But managers in utility companies understand that things aren’t that simple: every utility company is unique and the challenges are usually in the phase before any tooling can be selected.

Think of the data itself. It’s often stored in a variety of systems (perhaps including some that should have been shut down years ago). This is particularly true if a utility company has integrated several smaller suppliers without properly integrating all the IT systems. Understanding where all the data is and how to open it up for analytics can be a very tough first step. No standard data science tool will solve that problem.

Also, utility companies do not want to become dependent on a specific supplier. They want to understand analytics themselves and find out how to use it before selecting any specific software solution.

Data science comes with its own challenges: good data scientists with a good grasp of the utility sector are hard to find. This is where companies like VORtech can make a difference. Our data scientists and consultants can help utility companies with understanding their data-related challenges and provide solutions. By working together with our clients we make sure that the knowledge is transferred and by using open source tools such as Python and R we ensure independence of any supplier.

In the foreseeable future, IT challenges of the utility sector are becoming increasingly data-oriented. The potential benefits are huge for those companies that manage to integrate data science properly into their operational processes.