DataOps predictions for 2025 - N°2: From AI Hype to Data Fundamentals
Back to our DataOps predictions. In this article we dive into what AI will bring in 2025.
Just before our winter break (we are based in Europe, it’s getting cold here) we published our first 2025 prediction: The Fog Clears, But Slowly. This prediction was all about the importance of DataOps, the increasing role for scale-ups active in this domain and the tough times traditional OT vendors will be facing.
This next prediction is all about one of that other buzzwords of 2024…
…Artificial Intelligence!
We believe that in 2025 we will finally step away from the buzz, and back to fixing our data fundamentals!
The cornerstones of each and every modeling problem (which includes the more advanced AI use cases) are qualitative data and domain expertise.
Domain Expertise
We’d like to quote Jon Weiss who rightfully stated in our recent podcast:
“AI cannot augment domain expertise”
However, we see more and more ways of how AI applications can learn from real world problems operations are facing, for example by incorporating knowledge graphs.
Qualitative Data
Regarding the data problem: we’ve seen more and more companies taking a few steps back. They go back to the drawing board and start designing a reliable and scalable data platform first. Sometimes because they tried some pilots and quickly understood scaling them proved to be a tremendous task. Others just gave up when they saw their data swamp…
Our Industrial Data Platform Capability Map is definitely the place to be if you want to understand how to develop a scalable architecture:
We believe 2025 will be the year where the dust settles and the real business cases start popping up.
But next to what we are saying, let’s also take a look at what others believe:
Deloitte: Manufacturing Industry Outlook
Digital and Data Foundation Investments: Manufacturers are prioritizing targeted investments in their digital and data infrastructures to boost innovation and tackle persistent skills gaps and supply chain issues.
Artificial Intelligence (AI) and Generative AI: The report highlights the importance of AI and generative AI in manufacturing, encouraging companies to focus on high-return-on-investment (ROI) applications. This includes leveraging AI for predictive maintenance, quality control, and supply chain optimization to enhance operational efficiency.
IT/OT Convergence: Integrating Information Technology (IT) and Operational Technology (OT) is essential for manufacturers aiming to create a cohesive digital ecosystem. This convergence facilitates real-time data sharing and analytics, leading to improved decision-making and operational performance.
DataOps and Data Management: Effective data management practices, including DataOps, are crucial for manufacturers to harness the full potential of their data. Implementing robust data governance and analytics frameworks enables companies to derive actionable insights, driving innovation and efficiency.
Gartner: Hype Cycle for AI
In summary the most notable things mentioned in their article:
GenAI has passed the Peak of Inflated Expectations (was about time 🙃)
AI Leaders must expand the focus of AI beyond technical conversations (thank you, and not just the leaders, every org 👍 )
Data Governance is one of the biggest hurdles (yes, and that includes data management, data quality etc... 🫡 )
The two biggest movers on this year’s Hype Cycle are AI engineering and knowledge graphs (Knowledge graphs are machine-readable representations of the physical and digital worlds. I do see a lot of potential here for Manufacturing)
Our focus should shift away from LLM-only to Applied AI
Generative AI (including Large Language models) are not per se the most interesting thing about AI, and definitely not in Manufacturing/Process.
This post from Rick Bullotta is worth taking a look at:
Extra:
See also this article from Colin Masson, Director of Research, ARC Advisory Group: “The AI Wars: Battlefronts, Breakthroughs, and the New Era of the Industrial AI (R)Evolution”. Definitely worth the read if you are working on AI in Manufacturing!
And just some common sense…
