Is there gold in them thar' suppliers? Proponents of supplier development programs believe so. They point to the oft-cited statistic that suppliers are responsible for 70 to 80 percent of an original equipment manufacturer's costs these days. That means, these proponents say, that much of the opportunity for cost reduction can be found in the shop floors and hallways of the OEM's suppliers.
For example, Dave Nelson, vice president for global purchasing at tier-one auto supplier Delphi and a long-time evangelist for supplier development, once estimated that the team of supplier development engineers he ran while he was vice president for worldwide supply management at heavy equipment OEM Deere & Co. uncovered $22 million in hard savings by working with suppliers to take costs out of their processes, all at a price tag to Deere of less than $7 million.
Supplier development advocates suggest that OEMs must be willing to shoulder at least part of the cost of these programs because their small and midsize suppliers will not be able to foot the bill on their own, lack the personnel or necessary skills, or simply will not be willing to change long-established processes without a push from a major customer. However, a case in point from the aerospace industry demonstrates not only that opportunities do exist for process-based cost savings at smaller suppliers but also that some of those suppliers are ahead of their OEM customers in identifying and profiting from those opportunities.
Focusing on Cycle Time
Jamie Yelle, president and general manager of Royell Manufacturing, knew he had a problem. Royell, based in Everett, Wash., just north of Seattle, is a manufacturer of precision-machined components and assemblies for the aerospace and commercial industries. While a small operation, with 32 employees and an annual turnover of about $7 million, the 30-year-old company nevertheless services some of the biggest names in the aircraft industry, making it an important link in the aerospace supply chain.
Trouble was, the company was taking too long to produce a particular machined part for an aircraft assembly. The triangular-shaped part, a strut fitting made out of the exceptionally strong 6Al-4V titanium, was highly complex, requiring surfacing over all its faces, and that kept the part on the company's Hitachi four-axis, horizontal mill for 18 hours, or about six hours more in cycle time than Royell had originally bid on the part. Consequently, the part was costing the company 35 percent more to make than they could charge the customer, a large manufacturer of aerospace components, under the multiyear contract for the part.
Looking back, Yelle realized that in bidding the part, his company had made certain assumptions that did not fully allow for the complexity of the parts configurations and surfacing. "We made a mistake on the bid," Yelle says candidly. The excess cycle time on the part meant that Royell was not covering the capital expense of the Hitachi mill, and the opportunity cost of the time-consuming process denied the company the ability to take on new, profitable work for that machine. Finally, the lack of any margin on the part reduced the resources that Royell had available to invest in new technologies and to grow the business.
Not that Royell was a stranger to lean. On the contrary, like many manufacturers in the hard-pressed aerospace supply chain, Royell had been pursuing lean manufacturing for a number of years. In particular, the company had focused its lean efforts, first, on addressing work instruction issues; second, on points-of-use issues, making sure that machinists had everything they needed at the spindle; and three, on simplifying the tooling concept and mistake-proofing the tooling process.
However, Royell had not previously addressed a part's cycle time on a machine. "Most of our other initiatives addressed strategies for the employees while they were working offline, away from the cycle time aspect of the part," Yelle says. "This was the first time we really approached optimizing the cutting cycle."
Getting Good Data
As a first step toward addressing the problem, Royell conducted an internal review of the part, producing several ideas to improve the manufacturing processes. But the company found that it could not come up with dollar figures for how much it would cost to implement the process changes and what the return on that investment would be. "It's always easier to make business decisions when you have good data," says Yelle, while adding, "In this case, we didn't know the implications of the recommendations."
In the meantime, Yelle had become aware of a startup solution provider called Vykor, based about an hour's drive south of Everett, in Renton, Wash. Describing itself as a provider of strategic sourcing solutions for engineered parts and tooling, Vykor uses a combination of technology and professional services to map, assess and quantify manufacturing processes and their associated costs, providing a client with a roadmap and a cost-benefit analysis for optimizing those processes. The provider's business model calls for it to work with the larger original equipment manufacturers that are willing and can afford to pay for the provider's services in assessing their suppliers' processes, with the idea that the OEMs will be able to share in any cost savings their suppliers experience. In Royell's case, Yelle was sufficiently intrigued with the idea of being able to see a cost-benefit analysis of a "to be" process that he was willing to assume the cost of Vykor's services. A couple months after first learning about the provider, Yelle brought Vykor in to do an assessment on the strut fitting.
Royell assembled the shop floor staff working on the part, including the machinist, the company's lone computer-aided manufacturing (CAM) programmer and the shop supervisor, along with Yelle, to meet with Vykor's professional services team. The manufacturer's staff walked through the current process for producing the part and also provided Vykor with a detailed description of the part, including such data as the engineering definition, part models and a manufacturing definition.
Dan Christiansen, vice president for professional services at Vykor, says this review gives the provider an understanding of the current process and an appreciation of the constraints under which the manufacturer is operating. "We do a very objective assessment of the shop to make sure we're not going to propose technologies and strategies that are outside the realm of their capabilities," Christiansen says. That includes exploring how open a manufacturer is to new technologies or non-standard processes to improve shop floor productivity, as well as learning about the manufacturer's current lean initiatives to ensure that a proposed new process jibes with those efforts. In Royell's case, for example, the company was interested in simplifying the tooling rather than creating all sorts of new tooling.
Vykor entered the collected data in its GPF Advisor solution, a process-mapping and design tool that produced a baseline "as is" process in the provider's "Global Production Format" (GPF). The provider's team then evaluated the current process using intelligence from academic research, cutting tool vendors, lean experts, machine tool manufacturers and Royell's own manufacturing experts, creating alternative part-production strategies in line with the manufacturer's objectives and capabilities. For example, to simplify the tooling process, Vykor's proposed strategies called for taking four or five tools out of a process that previously involved nine setup tools. Elsewhere, the provider recommended using a higher-grade carbide tool that was more expensive but that had significantly better performance.
To validate that the proposed strategies would be more efficient than the current process, Vykor uploaded the GPF file into its Cost Advisor solution, an online part-management application that details the costs, manufacturing strategies and run-times of engineered parts and tooling. The solution emulates the run-time of the manufacturing strategy captured in the GPF to produce both non-recurring and recurring costs in effect, it shows the cycle time and cost of the "to be" processes compared to the "as is" state.
With those results in hand, Vykor went back to Royell to present the proposed strategies, the changes that each required and the benefits. The provider believed that the new strategies could reduce the cycle time on the strut fitting by more than 60 percent, well over Royell's requirement to cut the cycle time by at least 35 percent to make the part profitable. Those figures impressed Yelle, but he was more impressed by the cost-benefit documentation that backed up the cycle time claims. "What we liked more than anything about the whole process was that it was scientifically based," he says of the provider's approach. "It was very easy for me to make a decision that we could increase our throughput on this spindle enormously through the cycle time reduction and increase our margin."
Putting a Thumbprint on the Strategy
Royell did not accept the new process across the board. Vykor had recommended a particular inserted cutter, for example, but Royell wanted to stick with a different cutter that it already had in its shop. The provider revised the strategy to include the existing cutter. That's not unusual in the give-and-take of finalizing the strategy, according to Christiansen. "Obviously they have to implement it, so they have their thumbprint on the strategy," he says.
Once Royell and Vykor had finalized the new process, the provider delivered a GPF file to the manufacturer with all the information necessary to program the part, including the tools, the use of the tools, the tooling approach, the machines and the tool-path strategy. Royell did the necessary programming in Mastercam, a computer-aided manufacturing application. With the first part that Royell cut using the new process, the company achieved 98 percent of the efficiency that Vykor had forecast, cutting the cycle time on the part by 61 percent.
The whole process had taken about two months, start to finish, and while he did not discuss the cost of engaging Vykor's services, Yelle says that Royell saw payback on its investment in about two months. The return on investment came in part through the reduced cycle time on the part, but also in the windows of spindle time that the new process opened up on Royell's equipment. "With the extra time freed up by the new, optimal manufacturing strategy, we were able to win a new contract and bring in additional revenue," Yelle explains. Royell also picked up bits of programming strategy that it will apply elsewhere in its operations, and the company learned about certain new cutters that it is now introducing for other parts. As for the strut fitting, Royell will continue to produce six to eight of the part for at least the remaining two years on the contract.
Chris Wright, production manager at that company, says that one of the key learnings from the implementation was that although Royell had been practicing lean manufacturing for a number of years, opportunities for improvement always exist. "I think the most important thing that this helped our employees to see is that the term 'best practices' only describes a snapshot in time, and implementing lean manufacturing is continuous," Wright says. "It showed people that even though they thought the process was good enough, it really wasn't. It stretched people to acknowledge that there are always ways to improve, even in cycle time. It demonstrated that you have to be continually working and looking for new ideas and approaches."
The lesson here for OEMs is twofold. First, as Royell demonstrated, opportunities for significant increases in efficiency and reductions in production costs exist on the shop floor, waiting to be uncovered. And second, many smart suppliers are moving ahead with their own initiatives to take advantage of the benefits to be reaped from better processes, taking the initiative and a potential share of the cost savings out of the hands of their OEM customers. There is gold in them thar' suppliers. It's just a matter of who's going to strike a claim first.
