Question: What role do you view big data analytics and AI playing in the future of lean manufacturing?
Answer: Wow, this question came out of the sky. Not in my wheelhouse. My initial reaction was, “How in the world would I know? I have enough trouble keeping up with the lingo, e.g., IoT, big data analytics, AI (artificial intelligence) and a partridge in a pear tree!” But then I reached out to a good friend in Chicago, John Shelby, who recently completed his master's degree in analytics at Notre Dame. I asked if he could educate me in plain English on the question posed. He shot me a link to a McKinsey study in which they are tracking their 2011 forecast of how these evolving technologies would impact retail, supply chains, healthcare and, yes, manufacturing.
John’s input helped me understand the difference between “lights out” manufacturing operations versus AI. In the case of “lights out,” it is an advanced state of operations where machines run with no human interaction, i.e. “robots.” This has been done modestly in factories the last several years but has not yet been widely adopted. The machines are programmed by humans and are still subject to human inefficiencies.
In some forms of AI, on the other hand, machines can actually teach themselves how to optimize their performance as they can run through various scenarios at lightning speeds, identify the best processes and train themselves to achieve a desired outcome. Talk about a brain stretch! I think it’s likely that AI will ultimately be the real game changer in the factories once new ERP platforms are in place, but this may still be a decade or more away. But imagine manufacturing companies that can absolutely guarantee 6 sigma, every time, on the customers’ critical-to-quality requirements. If your company is among the leaders in achieving this kind of breakthrough, what would that be worth to your company in terms of increased market share, low costs, perfect service, operating margin, shareholder value?
This all sounds like Star Wars hype to me. This scenario is likely a decade or more away. What I want to know is this: Can any of this technology be applied where I’m most comfortable, i.e., in a factory, on the shop floor? What can manufacturing leaders do with this thinking and application on a practical basis?
As John and I had a back and forth on these topics, he pointed out that big data analytics has been around for a decade or more and, while it continues to expand, storage is cheap and readily available and should represent no barrier to moving forward. That’s when my lightbulb started to come on more brightly, and it hit me that the issues for big data and IoT are more along the lines of outdated thinking in the programming of ERP systems.
Corporations are still spending untold millions of dollars buying “new” ERP systems that still rely mostly on traditional thinking and reporting. Further, report formatting, because of the same outdated thinking, isn’t structured in a way to get actionable, real-time reporting. Why can’t the ERP system simply be linked to the machine’s PLC and provide data in real time? Or must we always have a secondary system, often not integrated with anything else, to have what shop floor people need to more effectively manage the business minute by minute, hour by hour?
Historical reporting is of no value on the shop floor. We can’t fix yesterday or last week or last month -- or even the last hour. We need to know if we’re making the numbers in real time. I wish I could look around corners and know if this is about to change. What will the next generation of ERP systems design look and act like?
It is a real brain stretch trying to think about the possibilities of how we might use these developing technologies in the coming years of manufacturing. These developing technologies and their implementation will be the next big paradigm shift that will cause manufacturing companies to either take a huge leap forward or else be left in the competitions’ dust.
I hope our readers find this prospect as thought-provoking as I have. It starts a conversation in each company whose time has come and, yes, it’s a bit overwhelming to contemplate right now. But now is the time for all C-Suite leaders to follow these technologies carefully, partner up with thought leaders in this area, and strategize how to be among the companies who get up to speed early and commit to the new technology -- in fact, even help develop it. Some recognizable manufacturing names like Rockwell, GE and Siemens are already out front. Frankly, our software friends need manufacturing practitioners’ input desperately to maximize the potential of the technology. Throwing the next release over the wall without our input would be a serious mistake. But aren’t the prospects exciting?
Interestingly, I just returned from the AME (Association for Manufacturing Excellence) International Conference in Boston and had a wonderful chat with a representative of a company called MCA Connect. I was attracted to their booth when I saw a bullet point that said: The “Voice of the Operator.” They are partnering with Microsoft and designing a system that enables easy data migration, i.e., that allows you to pull in traditional manufacturing data from any ERP system, convert it into lean manufacturing data and assign it to the right fields in their “Dynamics 365 for Finance and Operations,” built with lean manufacturers in mind.
I asked a couple of penetrating questions that hadn’t yet been considered in the system design (reread comment above about seeking practitioner input). In fact, the system they showed me had been hurried along so they could introduce it at the AME conference. But it has great promise. I asked to see scrap information, OEE, etc., and got a clean dashboard of color-coded metrics. I looked at some causes of availability problems in the OEE data and saw specific maintenance issues that had been identified. Tremendous potential. Oh yes, and we need to add one more acronym to our vocabulary, i.e., IIoT, Industrial Internet of Things.
My excitement calmed when reality set in and I realized this: While the potential is enormous, it will only be fully realized when the software writers of ERP systems catch up with current thinking. Why? Because using this slick new access tool to tap into the wrongly designed ERP system data will simply give manufacturing the same wrong information at the speed of light. ERP programs must be redesigned and cleaned up in concert with these new access systems. In fact, one might wonder why access programs would even be required if the ERP systems provided the same capability? In any case, until ERP programs smash long-standing paradigms and catch up with state-of-the-art manufacturing needs, they will continue to represent a formidable constraint to achieving the full impact of the improvement potential that is now obviously possible.
Yeah, I know, it still sounds like a Star Wars kind of fiction. But here is the information I picked up at the MCA Connect booth:
- 79% of companies have started an IoT initiative.
- 44% of manufacturers have a defined digital strategy.
- 47% believe their business model will be obsolete in three years.
- 89% see customer experience as a basis of competition.
I wonder how many of us will be ready when the Star Wars world becomes reality in the next few years. At first, just thinking about this gave me a headache. But the more I thought about the possibilities, I was overcome with excitement and optimism that manufacturing is on the cusp of a new age. We’ll all be wise to sign up and help make it happen.
"Those that say it can't be done need to get out of the way of the people who are already doing it." - Joel Barker, futurist and author
Larry Fast is founder and president of Pathways to Manufacturing Excellence and a veteran of 35 years in the wire and cable industry. He is the author of The 12 Principles of Manufacturing Excellence: A Leader's Guide to Achieving and Sustaining Excellence.