top of page
Search
  • Writer's picturePaul Hogendoorn

AI for Manufacturing - the Good, the Bad and the "Meh"




AI has arrived. Everywhere you look, everything you read, and everything you hear, AI is said to be the big “what’s next”. The truth is, it has been slowly creeping into our lives for years, but it wasn’t until the “whiz-bang” emergence of ChatGPT that it caught mainstream attention. Self driving cars, autonomous industrial vehicles, all sorts of financial systems, advanced video games, social media and automated marketing systems are all examples of AI in use today.


As a quick summary, Artificial intelligence (AI) technology aims to simulate human intelligence and problem-solving tasks by rationalizing data and taking action to achieve a specific goal. As far as manufacturing is concerned, there are good applications for AI, bad applications, and applications that make little difference.


First, lets look at what I consider as bad, or potentially bad, applications for AI.


Relationship management and relationship building are less than ideal applications for AI. Having a robot pick up a conversation with me when I browse a website is OK, but, if I’m seriously considering purchasing someone’s product, my decision to choose one vendor over another comes down to only two or three things: does the product do what I need it to do, what is the cost, and, what level of comfort do I have in that company. Any time the goal is to create relationship, or build relationship with customers, or even just with other people, abdicating those tasks (and in effect, responsibility) is doing exactly the opposite. Even with the writing of this piece – an opinion piece reflecting my personal opinions – the goal is to be genuine. Delegating the task of writing it to AI quickly becomes ingenuine – its no longer my personal, human opinion, it’s artificially created. Opinions need to be genuine, and so do relationships.


Real human relationships require real human interaction - not a mimicked human interaction. As more companies delegate more of these tasks to AI, it provides a growing differentiating opportunity for others that choose to keep their relationship building efforts genuinely personal. (Take note of the two major banks and how much they are now spending on marketing trying to differentiate themselves as being “more personal”). Don’t delegate your genuine relationships to technology; they are too valuable (and remedial advertising too costly).

   

Next, let’s look at the “meh” (shoulder shrug), low value applications for AI.


In my opinion, anything that is backwards looking is not without any value, its just of far less value relative to the power of AI. To derive any benefit from historic data means coming up with a plan of action to improve on the analyzed historic results, putting that plan into action, and then monitoring results for a period of time to determine the effectiveness of the action. The whole process is not without value, its just (in my opinion) not an example of an application that leverages the power and potential benefit of AI technology.


AI, in my opinion, has far more potential benefit for manufacturers looking forward, by optimizing product design, production routing, machine and human resources scheduling, than it has looking backwards, trying to automate or suss out additional insights on what didn’t go right last week, last month or last year. Tool and die shops, fabrication shops, and steel service centers have been using software effectively for years that optimize machining time and material use on a machine by machine or sheet by sheet basis, but the entire job scheduling and routing is still generally done manually.


In the last few years, I’ve had the privilege of working with several thought leading manufacturers that figured out that improved planning yielded the biggest overall efficiency results than simply looking backwards and trying to measure and improve on their OEE, with several of them doubling their capacity by eliminating wasted time and energy through better planning – in other words, looking forwards, rather than always looking backwards. They gave their people better tools for planning and not just better tools for measuring and reporting. Its like giving your people a radar (or GPS) to look forward rather than just a mirror to look backwards.


The same holds true for product quality. As one anecdotal illustration puts it, a problem prevented in planning is a penny; a problem resolved in production a dime; solved on the showroom floor a dollar; and repaired on the customer’s driveway, ten dollars. With that in mind, why waste the immense power and potential of AI looking backwards for insights when its far better to be used looking forward for planning?


AI technology holds plenty of potential for manufacturers in North America. My hope is that its not just used to mimic, simulate, or replace human thought and action, but to aid, augment and advance human thought and action.      


32 views0 comments

Comments


bottom of page