The Digital Chemical Industry – Artificial Intelligence

Share

Of all the emerging digital technologies, Artificial Intelligence (AI) has possibly the most potential to assist chemical industry companies to achieve disruptive transformations both in terms of their performance and value generation. AI technologies are likely to have a huge impact on the value chains and profit pools within those chains that will have far-reaching implications for the structure of the entire chemicals sector. AI was identified as one of the critical digital technologies to be mastered in Europe in Cefic’s recent Mid-Century Vision document ‘Molecule Managers’. This article is the third in a series to highlight the possibilities of digital transformation and its role towards sustainable solutions.

The potential of AI technologies

The term AI is used to describe machines and processes that mimic ‘cognitive’ functions that humans normally associate with other human minds, such as ‘learning’ and ‘problem solving’.  “In terms of digital technologies, AI covers machine learning, reasoning, computer vision and speech recognition and autonomous or self-organised operations,” says Dr. Martin Winter, Innovation Manager at Cefic.

AI-technologies have enormous potential to be applied within early product and process development stages to significantly speed-up innovation, allowing a more efficient ‘idea-to-market’ process. For example, research productivity can be increased by AI-enabled access to all previous relevant results and domain knowledge during the experimental design phase.

AI will allow a quicker response to market and customers’ demands through integration of customer requirements into the R&D-processes. Digitalisation will enable the ability to integrate lifecycle thinking and advanced sustainability assessments to achieve the targeted solution.

AI can enable optimisation in all stages of the value chain and provide the objective base for value extraction. Data truly becomes the capital for reliable and precise models as the basis of predictive manufacturing and enhanced industrial competitiveness.  Data can flow freely from the time that the production processes are given a design and dimensions, to the installation and start-up phases and finally to operation and maintenance. The Physical Factory that manufactures the product is closely linked to a Digital Twin that is continuously using the same data flows to simulate operations to further optimise processes and material and energy flows.

For the European chemical industry, AI can make industrial processes safer and cleaner and help invent new molecules for specific customer needs, including increased circularity of materials.

What to expect long-term

The long-term goal would be to develop AI machines that offer suggestions for optimisation autonomously. The expectation is that machines will learn from observing operator interactions with process control systems and historical data. “Humans are normally only able to think and react linearly. They are able to learn and build on existing knowledge for decision making but usually lack the capability of true multidimensional thinking,” explains Winter. “Machines could be capable of utilising all of the available data and ‘learnings’ to help us humans make better decisions.” AI should enhance human reasoning and decision making rather than replace it.

“Overall, AI promises to relieve humans from routine tasks and to promote better decision-making in the industry to achieve disruptive transformations that strengthen global competitiveness as well as contributing to a carbon-neutral economy and to circular-economy solutions,” claims Winter.

AI investments: economic and human

Investment is already huge. The Cefic ‘Molecule Managers’ document quotes a European Commission estimate that private industry invested more than €30 billion in AI globally in 2016 alone. AI technologies are developing very fast, but adoption in the chemical industry is often challenging, meaning support for research and innovation actions to speed-up their adoption of AI-technologies is needed. The new SusChem Strategic Innovation and Research Agenda (SIRA), to be published in November, will elaborate a portfolio of detailed research and innovation priorities in AI to support a sustainable chemical sector.

Specific AI solutions for the sector will create a large demand for highly digitally skilled human operators to develop, monitor and manage AI-based operations. “Investments in digital education and skills is therefore also a very important priority,” concludes Winter.

Over the next few years it will be critical for the chemical sector to find the right applications of AI technologies that can bring the most benefits in terms of reducing energy consumption, reduction of environmental pollution, realisation of a full circular economy, cost reductions and product quality improvements.