By now, we are all used to being surrounded by “smart” machines — smartphones, smart televisions, all sorts of smart appliances at home and smart instruments in our production sites. We are also familiar with connecting our smart devices and sending and receiving information from just about anywhere. And the amount of information that we can gather has increased astronomically, leading to terms like “big data” and to a focus on data analytics in order to put the vast amounts of available information to good use.
Machine learning and artificial intelligence
Alongside this boom in digitalization, more advanced computational techniques and related technologies have given rise to machine learning, where machines do more than just gather and share information, but actually learn through automated data analysis and by identifying patterns. Machines can then predict what will happen and offer recommendations to prevent negative occurrences. One practical application of this is predictive maintenance. When given even broader data resources to process, machines can learn more complex associations, for example facial recognition, in a process termed deep learning.
Machine learning is part of artificial intelligence (AI), which encompasses technologies that allow machines to learn and perform functions that mimic human intelligence. Rapid advances in computer software and storage capabilities, along with the massive increase in data availability through smart devices are enabling broader applications of AI.
Practical AI
Industries are increasingly incorporating AI into their businesses. In a recent survey by McKinsey & Company [1], half of the 2,395 participants from a wide range of industries said that their companies have implemented AI in at least one function. The chemical process industries (CPI) are among the adopters of AI. In April, Evonik Industries AG (www.evonik.com) became the world’s first chemical company to participate at the MIT-IBM Watson AI Lab (mitibmwatsonailab.mit.edu). The AI Lab is a community of scientists from the Massachusetts Institute of Technology (MIT) and IBM Research who are working with businesses in researching AI. Evonik and IBM had already been investigating AI as partners for several years, and membership in the AI Lab extends this cooperation.
As another example, in December, Covestro AG (www.covestro.com) announced several pilot projects using AI. The company is, for instance, using digital technologies to optimize the manufacturing process for polyesters and to predict peak steam consumption at production plants.
In this issue
There are many practical applications for machine learning and AI being pursued by the CPI. To learn more, see our Newsfront “Artificial Intelligence: Advancing Applications in the CPI” on pp. 12–18.
And our two-part Cover Story on process analytical technology (PAT; pp. 24–34) explores exciting advances in process analyzer technologies. There is also much more in this issue and we hope you enjoy reading.■
Automation & Control
Intelligent machines
| By Dorothy Lozowski
By now, we are all used to being surrounded by “smart” machines — smartphones, smart televisions, all sorts of smart appliances at home and smart instruments in our production sites. We are also familiar with connecting our smart devices and sending and receiving information from just about anywhere. And the amount of information that we can gather has increased astronomically, leading to terms like “big data” and to a focus on data analytics in order to put the vast amounts of available information to good use.
Machine learning and artificial intelligence
Alongside this boom in digitalization, more advanced computational techniques and related technologies have given rise to machine learning, where machines do more than just gather and share information, but actually learn through automated data analysis and by identifying patterns. Machines can then predict what will happen and offer recommendations to prevent negative occurrences. One practical application of this is predictive maintenance. When given even broader data resources to process, machines can learn more complex associations, for example facial recognition, in a process termed deep learning.
Machine learning is part of artificial intelligence (AI), which encompasses technologies that allow machines to learn and perform functions that mimic human intelligence. Rapid advances in computer software and storage capabilities, along with the massive increase in data availability through smart devices are enabling broader applications of AI.
Practical AI
Industries are increasingly incorporating AI into their businesses. In a recent survey by McKinsey & Company [1], half of the 2,395 participants from a wide range of industries said that their companies have implemented AI in at least one function. The chemical process industries (CPI) are among the adopters of AI. In April, Evonik Industries AG (www.evonik.com) became the world’s first chemical company to participate at the MIT-IBM Watson AI Lab (mitibmwatsonailab.mit.edu). The AI Lab is a community of scientists from the Massachusetts Institute of Technology (MIT) and IBM Research who are working with businesses in researching AI. Evonik and IBM had already been investigating AI as partners for several years, and membership in the AI Lab extends this cooperation.
As another example, in December, Covestro AG (www.covestro.com) announced several pilot projects using AI. The company is, for instance, using digital technologies to optimize the manufacturing process for polyesters and to predict peak steam consumption at production plants.
In this issue
There are many practical applications for machine learning and AI being pursued by the CPI. To learn more, see our Newsfront “Artificial Intelligence: Advancing Applications in the CPI” on pp. 12–18.
And our two-part Cover Story on process analytical technology (PAT; pp. 24–34) explores exciting advances in process analyzer technologies. There is also much more in this issue and we hope you enjoy reading.■
Dorothy Lozowski
Dorothy Lozowski, Editorial Director