The Role of AI in Automation
An Industry Perspective from President Alex Pool
We have all seen the meme “Replacing automation engineers with AI will require customers to know what they want. We’re safe”. This humorous nod to the rapidly advancing world of AI sheds light on a classic problem within the industrial automation industry: scope definition. The industry has long grappled with its challenges to the extent that it has become an accepted norm. It doesn’t matter how you design and program your system – if you don’t know what it is supposed to do, you won’t succeed.
As coding languages have evolved, the programming environment and compilers have undergone significant changes to make the programmer’s role easier. Mundane tasks have been reduced, or even better, eliminated. We don’t need to remember line numbers anymore. We don’t have to recode the same divide-by-zero check every time we build a data input screen. We don’t have to fix the same bug in 95 out 200 valves across the plant since they are all derived from the same template. AI is an extension of programming growth. It allows programmers to focus on what the program’s core functionality and not the tactile things like syntax, structure, or readability.
The first language I ever programmed was LOGO. I was a young kid when I coded a simple list of commands that moved a triangle-shaped ‘turtle’ around the screen to draw line diagrams. After approximately 10 minutes of moving the turtle around the screen drawing random shapes, I asked the teacher what I should draw. Little did I know that in that moment, I was introduced to my first scope definition conversation. Things haven’t changed much in 40 years. We spend more time figuring out what the program needs to do rather than developing the program itself.
Regardless of the language, interface, and complexity, project requirements need to be clearly defined. Consider the best industrial automation programmer you’ve worked with – their work is accurate and fast and uses a deep toolbox of standard code. Artificial Intelligence (AI) now plays a similar role, but it does not eliminate the need for programmers who excel at problem-solving and defining the project scope.
AI isn’t revolutionizing our industry; it’s drawing attention to the hard part of figuring out the scope at the required level of detail required to program. You can’t program ‘TBD’.
What do I think the role of AI will be in industrial automation will be? It will alleviate the burden of redundant, repetitive, and monotonous coding tasks, allowing programmers to focus on the essence of industrial automation: figuring out the root cause of system challenges and problem-solving.