My wife and I are moving into a new house, and my wife has her heart set on a new refrigerator. It’s a modern appliance, but it’s styled like a 1950s refrigerator: rounded corners, levered handles, single door or double doors, and offered in a selection of solid retro colors, such as fire-engine red, lime green, or turquoise.
It’s an example of a concept that the manufacturing industry has been wrestling with for many years: mass customization. And it’s a far cry from the days when you could buy a Ford Model-T in any color you wanted as long as it was black.
In today’s global ecommerce marketplace, consumer expectations about personalization, performance, and cost are supersized. To deliver the flexibility and responsiveness mass customization demands, manufacturers must embrace emerging technologies such as the Internet of Things, artificial intelligence, and blockchain networked ledger systems.
The Only Constant
It’s obvious to anyone who’s paying attention that global supply chains are in flux thanks to new competitors, new regulations, shifting alliances, and skills shortages. Don’t expect those challenges to dissipate or sort themselves out any time soon. The opposite is more likely.
Manufacturers must understand that they and their supply chain partners must do a better job of exploiting the tremendous amount of information they’re generating in order to improve their processes and operations.
Unfortunately, it’s not possible for humans to react with the speed necessary to adapt to changing global requirements. Such dexterity is attainable, however, by leveraging emerging technologies.
Internet of Things
IoT isn’t new to manufacturing. But manufacturers’ use of it is still limited mostly to monitoring equipment and anticipating parts failures.
A more forward-looking manufacturer would benefit by incorporating its IoT equipment-testing results into a planning system to help predict points of failure in the manufacturing process. The company might then tie that failure data into its supply chain system so that it can execute time- and resource-sensitive processes accurately and automatically.
In a dynamic environment, an optimal failure rate is a moving target. A manufacturer that only recently considered a failure rate of 3% acceptable now finds that percentage unacceptable. Why? Because its costs are tighter, its customers are more demanding, and the consequences of failing to deliver on time are more severe.
Failure rate is tied to inventory, and inventory is critical to meeting customers’ expectations. To hedge against a projected 3% failure rate, the manufacturer over-orders by 3%. Which means that a lower failure rate cuts down over-ordering, and if the manufacturer can meet expectations with less inventory it reduces overall costs in many different ways. For instance, as the pace of product upgrades and changeovers continues to accelerate, holding less inventory reduces costs associated with obsolescence, such as discounting and storage.
When I was a planner at a high-tech manufacturing plant years ago, one of my biggest challenges was dealing with exception messages. An ERP or manufacturing system generates a tremendous amount of data as well as messages about potential conflicts represented in that data, such as missed start dates and stock outages. Many exceptions are minor, and the enterprise system itself handles some of them. But the hard truth is that no planner ever gets through all those messages and, therefore, the ability to prioritize exceptions is never as strategic as it should be.
Artificial intelligence in the form of machine learning and predictive analytics represents an opportunity for manufacturers to automate the prioritization of exception messages. These tools are able to analyze a constant flow of complicated data points to arrive at recommendations or even initiate a series of actions based on trends discovered in the data. Planners can then be confident that critical manufacturing processes won’t be disrupted by unacknowledged exceptions. Oracle is working with a manufacturer that’s automating the handling of its exception messaging using RPA (robotic process automation).
Like IoT, AI isn’t new to manufacturing. But for most companies, AI was too expensive and complicated to use in day-to-day manufacturing operations. Now both IoT and AI systems are offered as flexible cloud-based services. AI capabilities are even being bundled into enterprise applications. All of which makes using these tools much easier and more cost-effective.
Blockchain, most often associated with the Bitcoin cryptocurrency, can be used whenever there’s a need for a trusted, secure distributed ledger. One supply chain example is the use of blockchain to track and control so-called “conflict minerals,” used in the manufacture of automobiles.
A blockchain network’s transaction transparency gives manufacturers greater visibility into their partners’ practices and performance. The immutability of blockchain data improves order accuracy and delivery times as well as the ability to track delayed or lost goods. Blockchain’s efficiency in settling transactions and dealing with government regulations improves the overall performance of an extended supply chain.
Oracle is working with a manufacturer that’s using 3D printing to produce body and cab frames for autonomous electric vehicles. The company wants to leverage blockchain’s distributed network capability, but not in its manufacturing process—at least, not yet. Instead, the manufacturer wants to use blockchain if one of its vehicles is involved in an accident, to trace the origin of every part that played a role, in order to best determine the accident’s root cause.
The message of mass customization is this: If you’re not capable of upgrading your supply chain so that it works much faster, gives you much more accurate answers, and coordinates your supply and demand much better, you as a manufacturer won’t be around for much longer.
As for my new refrigerator, I’m hoping it can restock its shelves automatically. Given the way things are going, that might not be too far away.