Generative AI Improves Aker BP Data Management
/
From CIO
Cognite, which put together a definitive guide to generative AI for industry, hosted a webinar about how Aker BP uses generative AI in their data management practices.
The webinar was moderated by Gabe Prado of Cognite and the panelists were Øystein Aspøy from Aker BP and Redza Rosli from Cognite.
First, the panelists talked about data management challenges. Aspøy said, “We have data in different formats and data in different levels and of different ages. And so it makes it hard to go back and do everything from scratch in a newer format or according to today's specifications of an asset.”
The panelists discussed data trustworthiness and how hard it is to bring data together into a single source of truth, especially when data management processes are manual.
One big challenge is that data is often stored in documents—PDFs and images, especially. This format is not conducive to quick analysis like for data stored in a database or spreadsheet format. It’s a challenge to get the data extracted from the documents and into a more usable format.
Generative AI can be a solution to some of these manual data challenges. Aspøy specifically pointed out a couple of examples: Brownfield modification, supply chain and material master, and lifecycle information.
Prado pointed out that “When data becomes easy to access and you can trust it, multiple different stakeholders across the data value chain then have access to do their jobs better and with less manual effort.”
Cognite discussed their Document Parser tool, which takes unstructured data and turns it into structured data. This formatting allows for querying and connecting to other applications. Using generative AI in this process saves a ton of time and reduces the chance for human errors.
It’s important to the Cognite team to have trust behind these generative AI tools. Rosli said, “Between these processes we will have data quality checks and we're working on data quality features that will give more trust to the data or highlight to the user.”
As Aker BP is using these tools, Aspøy said, “We are able to make more data available for everyone in a faster and easier way.”
Cognite, Aker BP, and others are working on bringing generative AI into the data management process, but there's a long way to go. Rosli said, “The industrial data space is not as mature as maybe one would think it would be. So once you really drill into things there are still very outdated processes. So there's a lot of opportunity in that sense for reducing operational inefficiencies and really optimizing processes. So it's kind of exciting in that sense.”
Prado wrapped things up by saying, “If we zoom out a little bit, the other thing to unpack here is that generative AI requires an ecosystem of things to make it meaningful and make it work for industry.”
To watch the entire webinar on demand, visit the Cognite website.
Want to check out the latest tech, use cases and best practices for deploying AI with industrial/energy assets? Join us at energAIze: Energy AI Forum @ Industrial Immersive Week this March 5-7 in Houston, TX.
