The road to AI (artificial intelligence) in manufacturing is paved with, well, data, and of course opportunity. Many manufacturers recognize this and are beginning to take big steps forward. We see 64% of manufacturers are entering into an R&D or experimentation phase with AI, and 35% are deploying right now. Where are you on the journey?
In a recent conversation, Simon Floyd, GM, manufacturing and mobility, Microsoft, explains to me how quickly AI is being implemented and the steps companies can take to build an AI strategy in manufacturing. Let’s walk through the 7 steps right now.
Step 1: Focus on the business need. He says, “I have met so many manufacturers that build the beautiful machine for the sake of building the beautiful machine and then they wonder why it is not getting the adoption or it tends to fail.” As with any technology implementation, we must always start with the business need first.
Step 2: Build a good data estate foundation. He recommends, “Maybe you can start small and start with what does have good data quality.” Then, you need to have a plan for how to get there at some point. All organizations should be focused on creating a governance model for AI including having their own policies around the use of AI. They should be looking at the quality of their data and the types of data they wish to use.
Step 3: Invest in infrastructure for AI. Floyd adds, “The good thing is a lot of the infrastructure, and the data foundation is the same for what we need. Adding AI once you are there is a relatively simple task.” We need to make sure we build the foundation first, and then build out from there.
Step 4: Cultivate a culture for AI. He says, “There has to be a desire, in a way, to harvest its great qualities, and also what it can do for an organization.” As with any technology implementation, we need to ensure the workers are adopting the technology and that the organization’s culture comes around it to see growth and success with the technology.
Step 5: Create a strategy for AI. Floyd suggests, “Many organizations that don’t have an AI strategy; their AI strategy is to get some AI. It needs to be much more comprehensive than that.” The government has some good guiding documents.
Step 6: Foster skill development. He says, “Being able to reskill and get new AI skills is something I feel we all need in this industry.” In fact, he goes as far to question, “Are we skilled enough for what comes next, not just for what is now?” This is something I have covered in depth here, looking at what needs to happen to reskill and upskill in the manufacturing industry.
Step 7: Clean up data. Floyd recommends, “Think about having less fragmentation overall and getting more consolidation.” This is the final step in ensuring a successful AI implementation in the manufacturing industry.
Floyd believes AI will shine in manufacturing when it lets humans do what only humans can do, and AI will take care of the very complex computations that need to occur to optimize an entire process.
“Unto itself, AI is very good at understanding many different factors and bringing them all together in order to get optimization, thinking beyond just the machine itself and thinking more process wise is where I see AI going,” he explains.
If you want to learn more, make sure to stop by Microsoft’s booth at IMTS in Chicago from September 9-14. As we continue to move forward on our AI journey, let’s remember to first focus on the business need; recognize data quality and governance are key; and ensure the workers are skilled. We have a real opportunity to make big changes in manufacturing, if only we take the right steps to get there.
Want to tweet about this article? Use hashtags #IoT #manufacturing #artificialintelligence #Copilot #industrialtransformation #industrialmetaverse #MicrosoftAdvocate
Links for further learning: