DataOps is gaining traction due to its ability to help provide more powerful insights to everyone in an organization—not just CTOs and IT teams. Recently, Geir Engdahl, co-founder and CTO of Cognite, and a former Google senior software engineer, explained why he believes adoption of DataOps has become a business imperative and not just a “nice to have”—especially in the heavy asset and manufacturing space.
How does Cognite differentiate itself in the market?
Contextualizing data and making it actionable sets us apart. At Cognite, we work with industrial asset operators to unlock the value of their data. By integrating digital solutions into their core processes, companies can maximize operational efficiencies and become more sustainable. Achieving this, however, relies on the ability to make industrial data trustworthy, reliable, and actionable for domain experts, data scientists, and business analysts alike. That’s what we do at Cognite. Our unique core technology addresses this specific, high-value industry challenge via our Industrial DataOps platform, Cognite Data Fusion. This AI-powered data contextualization platform transforms data trapped in silos, as well as hard-to-work-with datasets, into contextualized, connected, and unified data models. These data models enable data consumers at all ability levels to easily build digital solutions.
What do you mean by “Industrial DataOps” and in what way, if any, is it different from traditional DataOps?
Industrial DataOps breaks down silos and optimizes the broad availability and usability of industrial data generated in asset-heavy industries, such as oil and gas, renewable energy, power, and utilities, automotive, and manufacturing. All these industries generate vast amounts of data and will reap greater value if they accelerate their transformation efforts. An Industrial DataOps platform provides real-time data support for its operational technology, as well as deep support for engineering technology, including drawings and CAD models. This is the key differentiating factor for DataOps in industry.
However, Industrial DataOps isn’t about technology alone. Collaboration between domain experts is essential in making data valuable and useful for data consumers across an organization. DataOps is a practice, a way of engaging across the organization to both share and reap greater value from data.
Data is only as valuable as the analytics behind it and the scale of people who use it. The convergence of data and analytics has made Industrial DataOps an operational necessity, but the data requires context. By automating the data process and creating one central, contextualized source of truth, we can ensure live data triumphs over its predecessors in the decision-making process.
What is Cognite able to provide that specifically handles the issues of data management in the area of heavy industry?
Cognite is the perfect marriage between digital and domain. We have the in-house expertise to tackle the fundamental data issues faced by industry. For us, AI is a critical tool to achieve fact-driven decision making and efficient management of the data supporting it. This is the way to reduce human error and build trust in the data. At Cognite, we often speak about “data liberation” as a benefit of DataOps platforms. This liberation maximizes your data extraction capabilities and makes it easier to adapt to a DataOps way of work that integrates with your existing IT and operational technology architecture. It also limits the need to invest in additional systems integration and operational data sources.
We also support our customers in developing a strong data governance model for their IT, operational technology, and engineering technology data. This dictates how new data is connected and integrated into the overall data architecture, and it also helps them grow in data maturity by enabling them to serve a growing population of data and analytics business users. Once the data seeds are planted, it will spread through the organization, with the aim to flourish into a full-fledged digital operation.
How are the challenges, opportunities, and benefits in an industrial setting different from those of companies in other sectors?
The data isn’t the primary challenge industries face, it is how to best access it and contextualize it, so it is actionable. Operational technology data is the raw material that enables organizations to build more efficient, more resilient operations and improve employee productivity and customer satisfaction. This operational technology data is available in abundance, but industrial organizations struggle to generate value from their increasingly connected operations. According to the market intelligence firm IDC, only one in four organizations is analyzing and extracting value from data to a significant extent.
The breadth of legacy systems and complex data types that have accumulated over decades of operation can seem an almost insurmountable hurdle, as companies still want to reap the value and get the full lifetime out of their previous digital investments. Rather than throwing the data out and starting over, we want to attach meaning to this data and traverse the boundaries between the different systems, sources, and data types, bringing it together in one common language.
In an industrial setting, is the combination of people, processes, and technology easier or more difficult to roll out?
The definitions emerging about DataOps as an overall discipline have meaning across the business landscape, but they have particular importance for industry. Not only do the traditional data challenges felt across all organizations become much more significant in industrial sectors, both in terms of operational consequence and inherent data complexity, there are also unique challenges—the key challenge being legacy operations with deep domain experts and often very siloed data and even functions across the operation. So, yes, I would say the DataOps challenges are greater, and the opportunities even more so. If you manage to unravel that data complexity and bring the people on board, transforming your heavy equipment operators into data citizens even, then you can say you’ve truly succeeded in your transformation.
What types of technologies do you leverage to make the insights possible?
Industrial DataOps platforms offer the combination of data-driven statistical and physics-driven process modeling and simulation. While each approach has its pros and cons, often an machine learning model based on a hybrid of the two will provide the best results. These tools empower developers with workflows compatible with third-party AI tools and other necessary tools to develop, train, and manage hybrid machine learning models. This enables them to operationalize use-case-specific data subsets efficiently and at the desired scale. Industrial DataOps platforms also enable data users with low-code or no-code application development and model lifecycle management tools. This democratizes DataOps and facilitates a more collaborative working model where non-professional data users can perform data management tasks and develop advanced analytics independently within specified governance boundaries. This democratization of data helps store process knowledge and maintain technical continuity so that new engineers can quickly understand, manage, and enrich existing models.
What are some examples of benefits your customers have achieved through their use of Industrial DataOps?
Efficiency is a key benefit. Whether it’s about cost savings, waste reduction, or achieving sustainability targets, by introducing industrial DataOps across the operation, the primary outcome will be greater efficiency in a wide range of areas. We see this outcome in every industry, from oil and gas operators to renewable energy providers to power and utilities customers.
What’s next for Cognite?
We will continue building on what we’ve started, developing our product so that it becomes easier and quicker for industrial operators to use data to transform their operations. That means we need more industry data models available out-of-the-box for customers, and we need to deliver more flexible data modeling with powerful queries, including graph queries on industrial assets. We plan to further develop data lineage user interfaces, so that they support advanced flows and give confidence in the origins and processing behind each data point. Essentially, we will stay obsessed with developing our product and our solutions for industry, continually finding ways for them to accelerate their own digitalization and reap the benefits of their data.