This is the first in a planned series of blog posts that will examine the topic of artificial intelligence (AI) and its applications in industry.
First of all, this is not designed to be just another blog on AI. What I hope to do in this post is present a common understanding and definition of AI, particularly in terms of how it can be applied in asset-intensive organizations.
The challenge, however, is that AI may mean different things to different people in different contexts, so the concept can be hard to define. That’s because it isn’t really a technology in its own right at all. AI can best be described as collection of different technologies brought together to enable a system — a process, asset or machine — to act with intelligence.
If we extend this concept, then AI-enabled business applications serve the purpose of allowing a system to act with intelligence by helping it sense, comprehend, perform and learn. Training a system through machine learning or deep learning is a core part of what makes it intelligent — and can be incredibly powerful in optimizing performance, accuracy and quality.
Rather than focusing too much on dictionary definitions, it helps to think about AI in terms of what it enables a process or an asset or a machine or system to do. Machine learning is at the core of what makes a process, asset, machine or system intelligent. Being able to decide how to act by analyzing data, rather than through hard coding, differentiates AI from other forms of automation.
The “Constellation of AI”
As many enterprises embark on defining their AI programs and digitalization plans, a framework is needed to help capture the essence of the AI paradigm shift and the resulting transformation of all business processes within an organization — whether related to breakthrough innovation, everyday customer service, or enterprise productivity initiatives. Of the many frameworks that exist that try to define and explain AI, the one I find the most intuitive and logical is called the “Constellation of AI,” a paradigm introduced in the book Human + Machine: Reimagining Work in the Age of AI by Paul R. Daugherty and H. James Wilson.
Under this paradigm, enterprise AI can be viewed across three levels. At the first level, the enterprise defines its use cases and business applications (the why and the what) that leverage data to drive greater value to its stakeholders. At the second level, we look at the suite of AI capabilities that can be utilized to power the business application. And finally at the third (central) level, we look at the various machine learning methods that can be employed to deliver the pre-identified AI capability (the how).
As an example, one of the most compelling business applications of AI in asset-intensive industries is prescriptive maintenance. Using this framework, at the first level, these applications are getting more prevalent across industries that rely on systems to reduce unscheduled downtime, improve the asset lifespan and boost overall productivity.
At the second level, these business applications (or intelligent software agents) can leverage one or more AI capabilities to forecast when an asset needs to be serviced. And at the third level, these capabilities are based on a diverse array of machine learning methods, ranging from supervised learning using regression and/or neural network models to semi-supervised learning for pattern detection and other machine learning (ML) techniques.
Each AI use case and business application can be fundamentally deconstructed using this framework, thereby enabling enterprises to build a holistic AI program, to clearly analyze the business value for each AI initiative and to understand the baseline set of requirements necessary to invest and steer an AI program.
The Constellation of AI is powerful framework that enterprises can use to build out their AI program. The framework emphasizes the business value of the use case and applications, helps draw focus on the pre-requisites to the enabling AI capabilities and masks the complexity of the underlying AI and ML methods. With advancements that are accelerating the democratization of AI and ML, these technologies are becoming a critical pillar of an enterprise-level digital transformation.
AI/ML Adoption in Asset-Intensive Industries
While compelling use cases exist and there is measurable business value to be derived from enterprise AI adoption, asset-intensive, process-based industries are behind many other sectors in terms of AI adoption. One of the primary reasons for this is enterprise maturity.
This challenge can be characterized by the need for new skills and lack of quality data. A recent survey from Gartner indicates that 56% of enterprise leaders feel they need updated skills to accomplish AI-enabled tasks, and 34% of the respondents say that poor data quality is a key concern when implementing an AI project. (Given that understanding, there is still a lot of value in proceeding with AI for prescriptive maintenance because it also employs techniques to get strong results out of the current state of data in any asset.)
Another reason for the slow uptake of AI is an insufficient understanding of benefits and use cases, as 42% of Gartner respondents said they don’t fully understand the benefits of AI or the implied return on investment (ROI). Quantifying the benefits of AI projects poses a major challenge to enterprise leaders. By 2024, 50% of AI investments will be quantified and linked to specific key performance indicators to measure ROI.
So there’s still some work to do across the industry, but the starting point is developing a common understanding of what AI is and a simple framework that helps enterprises to visualize their AI projects. With this foundation in place, I will devote future blogs to exploring best practices in applying AI to drive the digitalization journey, as well as some of the common pitfalls and the key benefits for asset-intensive industries.
In the meantime, if you’d like to get more of a feel for how digital solutions can help unlock new levels of performance for your business, please read our recent executive brief, Next-Generation Operational Technologies Enable the Smart Enterprise in a Changing World.