Energy Analytics: Evolving Beyond Data Intelligence

Tom Watson, Director Infrastructure Operations, Williams

Tom Watson, Director Infrastructure Operations, Williams

Williams is a premier provider of large-scale infrastructure connecting U.S. natural gas and natural gas products to growing demand for cleaner fuel and feedstocks. When it comes to the geographic footprint, our facilities are spread across approximately 30 states. There are literally thousands of devices connected to our approximate 15,000 miles long pipeline network. This gives rise to a huge predicament of integrating the data present at all these locations and creating a single source of intelligence that will present a holistic picture of the operational data, to optimize the overall business. Williams took a systematic approach and after looking at the available vendors, chose Microsoft as it could provide us the necessary tools to garner data intelligence.

When we operationalized Office 365, we ended up implementing the technology infrastructure with a focus on a smaller subset of technology partners. And with Microsoft stepping up the power of their platform on the energy analytics side in the last five years, this has enabled us to provide advanced analytics and support the client organizations for not only their consolidation and reporting needs but also to simplify infrastructure and programming approaches for BI. Though we have collaborated with Microsoft on this journey to produce analytical insights, we are also considering other vendors that will broaden our sphere. An organization can perform energy analytics through a number of partners and different systematic approaches. Currently, there are cloud-based or on-premise solutions, but the next few years will witness the emergence of many innovative analytics vendors in the energy market.

"To find a suitable vendor, organizations need to consider the cost, analytical and operational capabilities, and time to deliver predictive analytics"

By collaborating with technology partners, organizations can also leverage many trends—which I would define as “Uber Trends”—and increase the efficiency of the delivered data intelligence. The first one is cloud which is very important as for an oil and energy firm there are a lot of sensors involved that contain disparate data, which needs to be combined together to deliver intelligence. The involvement of innumerable sensors makes security paramount as the data gathering process has to be secure and seamless. Further, the decision to migrate to technology ecosystems plays a key role as well. Organizations should leverage the aforementioned trends and form a partnership with a suitable vendor to deliver energy analytics. To find a suitable vendor, organizations need to consider the cost, analytical, and operational capabilities and time to deliver predictive analytics. They should also create an organizational layout of demands and needs, and make a judgment of which vendor will be adept at providing the capability to integrate disparate systems and data involved.

At Williams, we analyze a number of parameters before collaborating with a specific partner. We underline the consideration for BI and the implementation aspects prior to deployment. Then, we consider the first set of data that needs to be encompassed into our business analytics suite. We also ensure that the employees are working closely with the system to find out “how does the business contemplate data?” We have to prioritize the available dataset which needs to be fed first and derive intelligence from. The bottom line is to operationalize the business and provide organizational leaders with data that they can leverage to make intelligent decisions.

To devise a competent system and process, IT leaders need to showcase their capabilities in deriving intelligence from the available data. It’s hard to talk to people about a business case if you don’t have a demo or an approach document to show them. Once you come up with cross-system and functional capability, leaders can showcase it in different situations, which can go a long way to make customers want for more data intelligence. It’s necessary that the firms solve one problem at a time and not try to boil the ocean at once.