By Pablo Ríos, Business Manager for Manufacturing and Energy industries, Keepler Data Tech
In today’s digital age, data has become a key asset for businesses in almost every industry. And organisations are in possession of a vast amount of it; from human-generated information, machine-generated data, Internet of Things (IoT) devices, edge systems, and beyond. In fact, it’s estimated there will be 175 zettabytes of data in the global datasphere by 2025; that’s every person on the planet producing at least 1.7 MB of data every second for the next couple of years.
For energy firms, leveraging data is highly important when it comes to improving operations, services, and building greater resilience. However, there are two differing approaches to data – utilising data as products and approaching data with a project mindset. Like all companies, there is a clear need to adopt the most effective approach, but which is more likely to see energy firms make the most of their data investments?
Adopting an integrated and scalable approach to data
The current project approach to data undertaken by the vast majority of utilities firms has the potential to cause a number of drawbacks, including lack of integration, limited scalability, siloed departments, high cost, and limited agility.
To overcome these challenges, energy companies should consider utilising data products – in other words, adopting a more integrated and scalable approach to data, which encourages collaboration and innovation across the organisation. This approach involves building a data ecosystem in an evolutive way and “productivizing” the data to facilitate the consumption of business units and stakeholders.
A data ecosystem is a combination of infrastructure and business applications used to centralise, consolidate, and analyse information to improve the company’s bottom line. Within the energy sector, data ecosystems are focused on learning about and improving business and industrial processes to make them more efficient and offer customers better and more personalised services.
Adopting a more integrated and scalable approach to data involves building enterprise data architectures based on cloud-native services that allow for the creation of “liquid platforms” capable of scaling on multiple axes. This type of architecture is modular and can include new components or services to manage data flows based on business cases that need to be addressed. For instance, a real-world example of a data ecosystem is the Lake House architecture that comprises a data lake as the central piece and a series of data repositories (relational, time-series, data warehouses) that structure, process, and store the data efficiently, allowing it to respond to different consumers of data: developers, data scientists, business users, partners, customers, etc.
Overall, this approach helps “productize” an energy company’s data to facilitate the consumption of their business units and stakeholders by providing agility to the implementation and productization of business cases. The minimum underlying infrastructure for the data platform is provisioned to carry out the use case, but the platform can be scaled in a modular and flexible way. To deploy these platforms, the company’s technical profiles need to have the necessary capabilities, and business users must upskill themselves as users of the data platforms.
Stand out from competitors, while achieving emission reduction targets
Energy companies currently have a very important challenge with the goal of zero emissions by 2050. Many of them have set emission reduction targets of more than 50% by 2030 as well as the drastic reduction of basic raw materials such as water or minerals. These objectives will be accompanied by engineering projects that will generate new data. As such, a data ecosystem that consolidates this information from the outset will allow data-driven companies to differentiate themselves from the rest.
The bottom line is that energy firms must adopt a more integrated and scalable approach to data to build a data ecosystem in an evolutive way and productize their data. This approach enables energy companies to improve customer service, increase efficiency, enhance safety, make better decisions, and foster innovation, among other benefits.
Take international natural gas infrastructure company Enagás, for example. For them, the maintenance of pump equipment is one of the most important tasks in the efficient operation of their plants and it was implemented either on a routine basis or after the failure of equipment. Enagás wanted to change this procedure and anticipate possible failures by observing abnormal behaviours in the data. Through a machine learning model, this process evolved to a data-driven predictive maintenance system. While predictive maintenance may not sound all that innovative, the approach of developing a data product for this purpose, with worldwide scalability to all company plants, relates directly to boosting efficiency and cost savings.
With a model for every pump, Enagás’ technical team are now able to evaluate the functioning of each pump individually, reducing costs and intervening only when necessary. What’s more, thanks to an API that was also developed, users can perform multiple trainings/inferences faster, making this task more agile.
For energy companies with different plants around the world, such as Enagás, designing a solution locally that you can scale globally in a matter of days is a highly efficient way to implement improvements on a global scale. The crucial formula for success – and what sits at the core of our own approach – is to think big, start small, [mobilise] and scale fast.
With specialist knowledge, energy companies can deploy real use cases, enable data platforms, and provide the necessary capabilities to their internal talent to maintain and evolve these platforms and use cases. In doing so, utilities firms can not only differentiate themselves from the competition, but achieve those challenging emission reduction targets looming over all energy companies’ heads.
