Issue link: https://iconnect007.uberflip.com/i/1124788
JUNE 2019 I SMT007 MAGAZINE 17 end of life of a product; you see a square wave demand where the factory is absolutely flat out, and then the next week, it's zero. ERP is not the right tool to use anymore; it simply cannot work fast enough. But that doesn't mean ERP has to be replaced; it has to be augmented and have help. If you go onto Amazon or eBay, for example, you are buy - ing something from a seller who doesn't even own that product yet and is simply taking your order and passing it to a factory that directs you. That is the way people are ordering today, and this is a brilliant model because they never had to buy that. They don't have cash out, and they have no risk because they only get your money with markup and then pass it back. This is the future of B2B business, so how can we allow that business model without bank - rupting our manufacturing operations because they're boom and bust? This is a look ahead to what Industry 4.0 represents in its truest form. The ability for manufacturing and the supply chain to react to that level of volatility without resulting to the increase of stock, which would simply be moving the stock from one place to another, has no net effect. How do we make them truly responsive and keep their productivity? This is similar to "built to order" on a mass pro - duction basis. It has been impossible in the past, but this is where intelligence—artificial or otherwise—is needed within manufactur- ing. And the decisions that intelligence makes are based on the data. The data coming from the customer demand and makeup of the product; the understanding of the availabil- ity of materials at short notice and optimiza- tion of automated processes; and the ability of humans to adapt and work on something dif- ferent come together and are all linked with the data. The way that the linkage is done is through modern MES technology, so this is not the MES technology of the past where you get a plan and follow it step by step like a piece of clock- work machinery. This is adaptive planning, so we haven't done long-term simulations or plans. We are going to wing it day by day in an intelligent way. Matties: That is a new strategy, but it optimizes everything. I know you said you didn't like the word optimize necessarily, but that's what you're doing. Ford: And that is what people are doing right now. I went to one show in Sweden and jok- ingly, but not so jokingly, said, "You're prob- ably only getting about 30% productivity in your factory," and everybody looked at each other. After the show, people came up to me, saying, "That 30% is wrong; it's 10%." People are already fac- ing that reality. We're not saying that you have to make the plan up as you go along because that sounds a bit reckless. But what you're saying is you are already doing that. You're pretending to follow an ERP plan, but you're not. Let's optimize what you're doing right now in practice rather than trying to get this archaic infrastructure to predict or simu- late more, which is irrelevant given that the demand changes day by day. Matties: And if you have all of the data coming in, it's not a difficult task. Ford: Right, because people in manufacturing have a lot of experience—they know what to do in a particular set of circumstances—but what holds them back is not having informa- tion. Matties: AI is the place, and we're going to see where this flourishes. There's a big race between China right now and the U.S. for AI and China is quite vocal about it. Ford: We are on the cusp of creating true AI. IBM's Watson is an excellent example; it's sim- ilar to a car engine that could be put in many different cars. Let's get in the car that is man- ufacturing.