In the old days, when a turbine blade for a jet engine came off the assembly line, a grizzled old-timer might have to take a file to it to smooth a rough edge to make sure it met specifications, Danny Di Perna (pictured above) recalls. Some parts might have had to be scrapped. As senior vice president for engineering and operations at Pratt & Whitney, Di Perna has been in the engine business for a long time. With 5,000 parts and burning fuel at temperatures of thousands of degrees, they are machines of stunning complexity.
“We collect the data in real time. Every dimension is captured. Instead of waiting until afterward, we’re getting an in-process inspection.”
But today, as Di Perna and his team gear up Pratt & Whitney to cope with a huge increase in the number of engines it makes for Lockheed’s F-35 Joint Strike Fighter and for single-aisle commercial airlines like the Airbus A320, employees at the end of the line rarely have to fix a single part. Robots using special lights pick up the blades as they are being manufactured, and inspect them to see whether they match the specifications in a central computer system. If they don’t, they are altered or fixed on the spot.
“We collect the data in real time,” Di Perna explains. “Every dimension is captured. Instead of waiting until afterward, we’re getting an in-process inspection.” Thus, when parts reach the end of the process of being produced, they are perfect, Di Perna says.
Pratt, a unit of United Technologies located in East Hartford, Conn., is launching into a new era of manufacturing with gusto. Workers are organized in cells, not in traditional assembly lines. They look more like scientists than old-fashioned, blue-collar workers because much of their work consists of looking at computer screens to make sure that parts are being honed, shaped and coated within a narrow band of tolerances. In addition to its embrace of real-time data collection, software and advanced robotics, Pratt’s suppliers make 80 percent of its parts, but it has vertically integrated production of the remaining 20 percent. This is where it defines its competitive advantage against General Electric and other rivals.
To manage that critically important supply chain, Di Perna has built a Star Wars-like control room where deliveries from every supplier are monitored on large screens. It also has merged its manufacturing, engineering and supply-chain functions into one 10,000-person organization under Di Perna, rather than allowing them to remain siloed. “There is a lot of power in creating that structure,” Di Perna says, because, in part, it eliminates the problem of engineers designing things that cannot be manufactured.
“Agile. Smart. Advanced. Industry 4.0. The Internet of Things. Lean manufacturing on steroids. What they all refer to is a dramatic expansion in the capture of data, improved software that links machines and systems that have never before been linked.”
Altogether, Pratt is going to triple the number of engines it makes a year by 2020. Welcome to the Next Generation of American manufacturing. Different experts use different catchphrases to describe the transformation that is under way: Agile. Smart. Advanced. Industry 4.0. The Internet of Things. Lean manufacturing on steroids. What they all refer to is a dramatic expansion in the capture of data, improved software that links machines and systems that have never before been linked, advanced robotics, sensors, new materials and even the early stages of three-dimensional additive manufacturing. The combination of these techniques promises to improve productivity, reduce costs and cut waste, while allowing for much more flexibility and mass customization of products.
One key is more powerful software, the connective tissue that allows machines to change their functions and connect into a more coherent whole. “The cost of a programmable logic controller has dropped in the past 15 years to about 15 percent of what it used to cost,” says Hal Sirkin, a Chicago-based senior partner of the Boston Consulting Group and one of the nation’s foremost experts on manufacturing. “You can basically buy automation cheaper.” He is co-author of The U.S. Manufacturing Renaissance: How Shifting Global Economics Are Creating an American Comeback.
The combined power of the new technologies—and the need for higher-skilled labor to manage them—is just beginning to hit American manufacturers. It shapes up as the most sweeping change in American manufacturing since the auto industry began imitating Toyota Motors’ lean production system a quarter century ago. “The single biggest trend is that the pace of change is accelerating,” says Tom Comstock, vice president DELMIA strategy and digital offers and user experiences at Dassault Systémes, based in Long Beach, Calif., a leading provider of software for design and manufacturing.
“Manufacturing used to be a lagging set of industries. They didn’t care about being on the forefront of technology. Now, we’re seeing a lot more emphasis on improving the manufacturing and the systems involved.”
The aerospace industry may be the most sophisticated in its embrace of Next Generation manufacturing technologies, but the nation’s largest industry, the automotive sector, isn’t exactly asleep at the switch. Ford Motor, for example, is embracing Next Generation techniques to transform its global design and manufacturing operations. It won the “Manufacturer of the Year” award from the Manufacturing Leadership Council in 2013 for a system it developed with Siemens that allows Ford engineers to simulate the entire assembly process of vehicles at different plants. That capability assists in helping the engineers to understand what can be manufactured and what cannot be.
“The rollout of new systems that link each factory to global systems has been part and parcel of CEO Alan Mulally’s efforts to create “One Ford.”
The system, called IntoSite, also helps create more flexible production lines that build different models at the same time. IntoSite relies on Google Earth infrastructure and is a cloud-based, web application that allows users to zoom in on a specific plant and “see” what is happening throughout the factory.
But Ford isn’t stopping there, says John Fleming, vice president of global manufacturing and labor affairs in Dearborn, Mich. He estimates that the company assembles 25,000 vehicles a day at its plants around the world and that each one requires 3,000 parts arriving from different countries. Within many of those parts are sub-systems that suppliers already have put together. Much like assembling a jet engine, it is a task of extreme complexity. The rollout of new systems that link each factory to global systems has been part and parcel of CEO Alan Mulally’s efforts to create “One Ford,” rather than allowing its businesses in different regions to function with high degrees of autonomy.
“These advances are extremely enabling from a manufacturing perspective,” says Fleming. “When we look at our quality analysis and when we look at linking all of our factories with a factory information system, being able to bring all that together and to look at it globally helps us make decisions on a daily and weekly basis.”
Factory information systems allow plant managers to know exactly how many vehicles are being made and whether their quality is at acceptable levels. The same information is available to workers on the line. Ford is now linking different factory information systems from different plants into a bigger, global system. It is a powerful tool because it can help the company better maintain equipment, decide where to store spare parts and drive down overall lower costs by improving capacity utilization. Take the maintenance function alone. “We put in a standard maintenance operating system,” Fleming explains. “This is where collection and analysis of data is really important. It allows us to look at optimization. What are the pieces of equipment that are likely to require maintenance and are therefore the most at risk? A lot of this is about prevention and risk management.”
THE RISE OF ROBO-MANUFACTURING
The new tools also have become essential in planning how parts will reach each factory for assembly into vehicles. “At the end of the day, we still have to make things,” says Fleming. “An analysis of the data and how the networks work is important, but it comes down to getting the right material in the right place at the right time—every time. It’s that interface between equipment and our team members that finally makes it all work.”
“An analysis of the data and how the networks work is important, but it comes down to getting the right material in the right place at the right time—every time.”
More advanced robotics is another piece of the equation, because today’s robots can have multiple functions, rather than simply repeating one specific task. “They can join, they can weld, they can rivet, they can handle materials and they can do all those things in a certain cycle,” Fleming says. “They may pick up a part, put it in a place and pick up a welder and weld, then pick up a rivet gun and rivet, then apply an adhesive. All those things can be done with robots.”
The robots are linked to a network that defines how they interact with each other. The next challenge is improving the way robots engage with human workers. In recent years, most robots havehad to work behind a gate of some sort to avoid direct interaction with a human for safety reasons. But as robots improve and become safer, Ford says it has started to station robots side by side with its human workers.
Fleming declines to say how much money Ford is spending on the next generation of manufacturing, but it must easily represent hundreds of millions of dollars a year. The company obviously feels that there are clear payoffs available. One implication of the new manufacturing model is that companies can no longer afford the luxury of allowing their different arms to function semi-independently. Rather than coming up with an idea and “throwing it over the wall” to manufacturing, the new Holy Grail is that engineers and designers must know whether products they are creating actually can be manufactured by using computerized simulations that are part of “concurrent manufacturing.”
The way this concept works is that the Computer-Aided Design systems in engineering are linked to process planning systems and to manufacturing systems, so that the company can simulate making a product even while it is still being designed and engineered. One outcome is that new products or new iterations of existing products can be introduced much faster, and a greater degree of customization of products can be accommodated. But all of it hinges on breaking down internal, bureaucratic walls.
A similar recognition is that supply-chain managers cannot be one step removed from actual manufacturing, which is what Pratt & Whitney has discovered in East Hartford. There, on the second floor of its main headquarters building, Di Perna has created a war room where large monitors cover the walls with information about which suppliers have promised deliveries of parts on what dates. “We want to integrate data to look at problems in the supply chain before they become problems on our production floor,” Di Perna explains. “We cannot afford to have an interruption.” This is a particularly serious problem in aerospace, where lead times are often quite lengthy. “You order something today, and two years later it shows up,” he says.
“We want to integrate data to look at problems in the supply chain before they become problems on our production floor.”
The supply-chain specialists in the room are using Microsoft’s extended customer-relationship management software to examine whether every supplier has ordered the raw materials to make the product that Pratt is expecting and whether the supplier has the capacity to get the job done. This is Big Data at its best because Pratt has access to many of its suppliers’ own computer systems.
If a supplier has not ordered key materials or does not have an assembly line ready to go, Pratt knows it is at risk of suffering a disruption. If they suspect an “event” or disruption is about to occur, Pratt’s supply-chain mavens have two large monitors where they can schedule face-to-face consultations with suppliers. If reassuring answers are not forthcoming, the supply-chain specialists conduct an “escalation” until the potential problem reaches the top management levels of both Pratt and the supplier—in a hurry. “In a supply chain, there is data everywhere,” says Di Perna. “That’s a critical tool for us.”
THE ADVENT OF ADDITIVE MANUFACTURING
There are many other strands of the new generation of manufacturing that are emerging, including 3D printing. The technology behind this potentially revolutionary way of making parts by printing layer after layer of a given metal or plastic has existed for decades. However, it seems to have finally established a clear foothold, particularly in aerospace with General Electric, Pratt and Europe’s Airbus all using it. “Already, planes are flying with 3D-printed parts in them,” says Dassault’s Comstock, a veteran of 25 years of manufacturing experience.
Imagine that rather than taking a block of expensive titanium and scraping and carving a part out of it—resulting in much of the titanium being wasted—a 3D printer can put down layers of titanium powder and seal each one so that the object has the characteristics of regular titanium—without wasting any of the precious metal. “You also could build products you’ve never been able to build before because of the technological limitations of machine tools,” Comstock explains.
The ability for everything to communicate with everything else is another potential disruptor. “If the material you’re using to build something can talk to your systems and talk to your equipment, you can think about very different manufacturing models than you have today,” says Comstock. “This is an evolution from bar-code spanners to RFID tags to intelligent materials and machines in the future.”
“A big machine tool will talk to the system and say, ‘Hey, I need to be maintained. I’m beginning to go out of tolerance.’ ”
Machines will have “intelligence” thanks to a proliferation of different types of sensors. “A big machine tool,” he says, “will talk to the system and say, ‘Hey, I need to be maintained. I’m beginning to go out of tolerance.’ ” Managing such a world requires dramatically different computer systems because it demands multiple flows of data in different directions, rather than simply the flow of data from the assembly line to a centralized computer system. “If all the materials and all the machines in manufacturing are [communicating] and producing data, you’re talking about a model that will be difficult to manage with traditional systems,” says Comstock.
Clearly, billions of dollars are being spent on transforming American manufacturing and many more billions will be spent as different technologies hit the plant floor. The new face of manufacturing will be far more productive and efficient, while allowing for greater tailoring of products, a trend called “mass customization.” Will it be a monopoly of large companies? Not necessarily. Big companies can afford IT departments and the brightest experts and consultants, but their sheer scale makes it difficult to quickly apply dramatic new ideas. Smaller and medium-sized companies may be able to embrace some aspects of Next Generation manufacturing more rapidly, giving them a first-mover advantage.
“This is all about entrepreneurism,” says BCG’s Sirkin. “Small companies are figuring out how to do this and growing because [when they] figure it out, they can show others how to do it.” The bottom line? New technologies are transforming and improving manufacturing at companies of all sizes—making keeping up with the pace of change a competitive imperative.