In my previous blog, we discussed the various ways a machine can learn and briefly introduced the notion of “Industrial AI” to help understand how artificial intelligence is being applied in real-world industrial applications. In this installment, we will look at how Industrial AI is shaping the digitalization strategy of capital-intensive industries and explore the top use cases that are propelling AI adoption in the industrial sector.
Bringing Applied AI to the Industrial Sector
In recent years, there has been considerable investment on democratizing the access to AI technologies through various AI/machine learning (ML) platforms, frameworks and toolkits. This has indeed accelerated the enablement of AI-based use cases, but it has not necessarily translated to significant business value, especially in the industrial sector. To overcome this adoption hurdle, there needs to be increased emphasis on democratizing the application of AI to domain-specific industrial challenges with a focus on business outcomes.
The key to making AI work in real-world applications is getting the learning right — and more importantly, making it valuable and actionable in an industrial business context. Therefore, the development of AI-enabled applications needs to be purposefully guided by domain knowledge to derive real business value.
This is where we get to the paradigm of Industrial AI, which combines data science and AI with software and domain expertise to deliver comprehensive business outcomes for the specific business needs of capital-intensive industries.
To go a little deeper, Industrial AI can be defined as a systematic, collaborative and integrative discipline which focuses on developing, embedding and deploying various machine learning algorithms as fit-for-purpose, domain-specific industrial applications with sustainable business value — for capital-intensive, process industries.
At AspenTech, we combine state-of-the-art modeling, first-principle engineering, advanced AI/ML technology and a comprehensive portfolio of asset optimization solutions to enable a systematic Industrial AI methodology. This is what allows the seamless integration of computational models with physical systems — across the entire industrial asset lifecycle. The powerful paradigm of Industrial AI overcomes the maturity hurdles of new technologies, focuses in on real-world use cases and delivers measurable return on investment on AI-driven initiatives across the enterprise.
As an example, at AspenTech we are uniquely positioned to democratize AI-based solutions across capital-intensive organizations. Combining the first-principle engineering (physics and chemistry) at the heart of these highly complex assets with AI capabilities will transform how work is done and elevate the operational efficiencies that can be gained. Think of the physics and chemistry as the “infrastructure” for safe and efficient operations, while AI capabilities act as the “enablers” of semi-autonomous, intelligent processes.
In the process industries, real-life behaviors of complex interconnected assets, processes and systems are defined by the design characteristics and capacity (limits) of the asset, which are captured in the model of the asset dictated by the physics and chemistry of the process. The AI, like previous multivariable and adaptive control capabilities, is used to gain greater insights to operate the asset within the physics and chemistry of the process and the process design limitations.
Top Industrial AI Use Cases for Asset-Intensive Enterprises
In the recent Industrial AI Market Report 2020-2025 from IoT Analytics, the analyst team identified a total of 33 different use cases that employ AI tools and techniques on connected data sources and assets of industrial enterprises. This study estimates the global industrial AI market size will reach $72.5B by 2025, up from just over $11B in 2018.
From their results, these are the top 3 industrial AI use cases:
Predictive maintenance is the single largest use case for Industrial AI, estimated to make up over 24% of the total market in 2019. Predictive maintenance makes use of advanced analytics and machine learning to determine the condition of a single asset or an entire set of assets (e.g., a process plant). The business goal is to predict when maintenance should be performed.
Predictive maintenance usually combines various sensor readings (sometimes external data sources) and performs predictive analytics on thousands of logged events. To understand a real-life Industrial AI use case in action, you can watch these videos that show how Aspen Mtell® uses autonomous ML-based agents to predict equipment failures, detect deviations from normal behavior and prescribe detailed actions to mitigate or solve future problems.
Quality, reliability and assurance is the second-largest industrial AI use case category at 20.5%. One of the key challenges is to enable decision-makers to maximize the economics of business decisions by going beyond the equipment level and accurately predicting future asset performance of the whole system.
Process Optimization is perhaps the most obvious and compelling use case, but still one of the most difficult to implement. This involves multiple AI-based capabilities across the system: automating repeat human tasks, enabling real-time decisions across various applications, augmenting the asset lifecycle and optimizing the value chain across different business dimensions.
This use case employs advanced ML methods, including reinforcement learning and sophisticated deep learning neural networks, to infer information and intelligence from different data sources, assets and processes.
These industrial AI use cases are being steadily adopted by asset-intensive enterprises, at varying levels, and will significantly transform their business. In the next part of this AI blog series, we will look deeper into how business leaders are preparing their companies to implement AI programs, the business value drivers of those programs and the keys to success for such digital transformation initiatives.
Learn more about how next-generation digital solutions can help you achieve new levels of agility and efficiency in our recent executive brief, Next-Generation Operational Technologies Enable the Smart Enterprise in a Changing World.