Last week I walked around the International Manufacturing Technology Show with 89,000 other participants at the McCormick Center in Chicago. I went to see how AI was affecting manufacturing and it wasn't quite the way I was expecting.
The sheer scale of physical machinery is overwhelming: from elaborate manufacturing equipment and numerous robots, to water jets performing precision cuts on metal. You also notice the international aspects--the big booths belonged to German, Japanese and Taiwanese companies, not so much American. US leads the world in many things, but manufacturing technology is not one of them.
There was one exception--the three big cloud providers: Google, Microsoft and AWS (Amazon). All, especially Microsoft, have dedicated cloud services for manufacturing. All were touting their AI services for manufacturing from predictive maintenance to Edge AI to generative AI diagnosing machine failures. It was less clear if the audience was listening.
When I walked around a Health Tech show last year, one could fee the focus on data and the excitement about AI even just a few months after ChatGPT went public. At IMTS the feeling was different. The most exciting "new technology" was allowing supervisors to get stats and adjust the machines remotely from the beach, offering retrofits to old machines to allow it to happen. One exhibitor showed me an app where a line worker could show a video to a distant supervisor to get some help (they need a new app for that?). I suggested that soon the supervisor could be replaced by AI in the app and they gave me a funny look.
The most interesting use of AI came from a small Oregon-based company Machine Research, which uses AI for small manufacturers to create bids from proposals. It's a small operations--just seven employees, basically only two developers working on the cloud.
Like academia, the manufacturing industry is slow to change. Replacing old machines and products are expensive--you can't just do a software update. Both presidential candidates are pushing to increase manufacturing in the US but unless we can take advantage of our technical strengths, embracing the future and Industry 4.0, we'll find it hard to compete. Tariffs and financial incentives will only take us so far.
I find it hard to believe that this wave of generative AI will make a difference in the majority of manufacturing jobs for the same reason plumbers and electricians won't be replaced by AI soon. No matter how smart the home, there will always be the need for a human to come in and fix things. Some jobs simply require skills orthogonal to those offered by AI. In general, AI is not reliable enough to get humans completely out of the loop, and the line of research on automation has been around for a long time and recent advances are incremental, not groundbreaking.
ReplyDelete