Viewpoints: Why Don't We Automate Military Airframe Assembly?
It's not uncommon to see tooling repeatedly positioning automotive components of the same weight and similar shapes as airframe parts to +/- 0.005 or 0.010" (+/- 0.13 or 0.25 mm). In addition, the Outer Mold Line (OML) or skin smoothness, and conformance to engineering design, outer skin alignment and fit, and paint quality meet—and can exceed—current military standards.
So why hasn't automotive assembly automation been widely applied to the manufacture of military airframes? The mantra of opponents—heard whenever automotive models for automation in military airframe assembly are mentioned—is "We don't produce the quantities of product that they do in automotive. We can't attain the ROI required to justify automation's widespread application."
In fact, examining an automotive model for capitalizing an automated production line shows that ROI is insensitive to units produced, and has more to do with return on revenue per dollar of capital expended. A good example is the Mini manufacturing facility in Oxford, UK. BMW produces more than 200,000 Minis/year.This volume represents a unit-average revenue stream of approximately $14,000 US, or an annual revenue stream of $2.8 billion (200,000 x $14,000). BMW invested over $1 billion in the assembly line, paint facilities, 750 robots, setup, and integration. They replace their robots every seven years, and anticipated model life at peak for the line is ten years. An average expectation for capital utilization is 8.5 years. Therefore, the anticipated revenue stream from the investment is $2.8 billion x 8.5 years, or $23.8 billion. Divide this by the capital expended to produce the revenue ($23.8 billion/$1 billion) and the return on investment is $23.8/dollar of capital invested.
The "Mini Return on Revenue" model can be used as a calculator for the return on automation of the modern military fighter/attack airframe. Let's say the three major defense manufacturers cooperatively assemble the airframe, and they share equally in its production.Then the proportional revenue stream for each of the partners is 1/3 of $35 million per airframe, or $11.667 million. And if, at peak, the estimated production rate is 200 units/year, with an anticipated life of 25 years, then the total revenue stream over the life of the assembly line would be 200 x 11.667 x 25, or approximately $58 billion.
It can be argued that each of the partners will not need 750 robots or the facilities that a Mini factory needs to produce 200,000 automobiles, so the airframe assemblers may not invest as much. But let's use the same capital expenditure ($1 billion) for an apples-to-apples comparison. The return on revenue for automating the airframe assembly would then be $58/dollar of capital spent, or a return approximately 2.5x the return that automobile factories regard as warranted.
The precision is present and proven; the return on capital investment is more than double. Why are we still hand-building military airframes with only token islands of automation?
The customer pays for the hand-built method, which uses huge tooling structures surrounded by workstands that prohibit and restrict access, slow assembly, and require hundreds of crane moves to transport each subassembly to the next subassembly. The design and use of even isolated islands of automation, much less a sophisticated automated assembly line, is (or would be) paid by the company.
Gee! The customer will pay the company $1 billion to make the tooling necessary to assemble hand-built airframes, or the company will spend $1 billion out of margin to automate. And because the product is less expensive to produce with automation, and the company is required to pass the savings on to the customer, which does it choose?
What is to be done? Unless the government looks to the manufacturing process, and adopts ownership and payment processes to recognize automation in the same way it pays for tooling for hand assembly, then there is no incentive (on the part of the companies) for change— other than token efforts.
It really isn't about how many units a factory assembles, or precision, or different product types, or any other reason. It's about who pays, and therefore the incentive for change.
Unless that changes, 20 years from now a visitor to any of the remaining military airframe assemblers will see what they see today: a complex, sophisticated product made of exotic materials, with thousands of hours of effort invested in it, being gingerly hoisted from one big steel box to another by a crane. And once positioned inside the box, the visitor will still see highly skilled mechanics perform the miracle of assembly—by hand.
This article was first published in the July 2006 edition of Manufacturing Engineering magazine.