In the circus, flexibility is a key attribute. Acrobats, jugglers, aerialists and other performers have to twist and turn innumerable ways. And, the people behind the scenes often have to jump through hoops and bend over backwards to ensure that the show goes smoothly.

In the three-ring world of manufacturing, machine builders and systems integrators have long touted the benefits of flexible, modular assembly systems. In theory, manufacturers benefit because the equipment can be used for more than one product line or more than one assembly process.

Today, being flexible is more important than ever. Agile assembly systems allow manufacturers to react quickly to shifting customer demands and shorter product life cycles. They can produce a wider variety of products; alter the mix of options or features; add new processes or assembly stations easily; and change volumes with minimal investment and changeover loss. Flexible assembly also allows manufacturers to implement production closer to a desired product launch date, with lower changeover costs and shorter timelines for those changeovers.

"Flexible automation systems allow manufacturers to be more competitive, because they can use one piece of capital equipment to manufacture many things," explains Mike Cybulski, vice president of Eastern operations at ATS Automation Tooling Systems Inc. (Cambridge, ON). "Using a common assembly platform reduces the need for spare parts."

Automakers are leading the flexible charge. They are scrambling to reduce the number of vehicle architectures they use globally. With flexible manufacturing, more than 75 percent of capital-intense tools and facilities can be reused.

"The automotive industry has experienced success with flexible automation because they are driven by short product life cycles or design revisions, and suppliers are required to react quickly," says Howard Speiden, customer dialog section head at Sortimat Technology (Schaumburg, IL).

In a flexible environment, vehicles are put together the same way, with shared program engineering and shared components. Multiple vehicle segments-cars, SUVs, vans, trucks or crossover vehicles-can be built off the same architecture. Variability in body styles and sizes, chassis sizes and other differences are easily accommodated. That translates into tremendous cost savings and reduced downtime during product changeover.

Flexible assembly has also been successfully adopted by consumer electronic and medical device manufacturers. "Both of these industries turn over their designs [with new and improved models] after a relatively short product life," says Ed Zimmerman, P.E., director of technical operations at Apex Automation Inc. (Elizabethtown, PA).

Any product that contains subassemblies that require the same basic parts and processes as other subassemblies, such as cell phones or electric motors, generally lends itself to flexible systems. For instance, in the electronics field, printed circuit board assembly machines are normally easily reconfigured. However, Zimmerman says other electronic subassemblies, such as connectors, may not lend themselves to flexible assembly.

No matter what type of product is being assembled, the ultimate goals of flexible assembly are the same: Reduced change-over time, greater responsiveness to customer demands, higher productivity and quicker time to market. But, exactly how much reconfiguring and reuse can be done, and how much it costs, is subject to debate. And, once products and processes become obsolete, there are questions about what happens to old equipment.

Multiple Definitions

A flexible assembly system typically consists of a standard platform and a series of interchangeable process and test modules that can be removed, modified and replaced quickly and easily. It allows for the future assembly of similar, possibly unrelated or currently undefined products.

According to Webster's Dictionary, "flexible" is defined as "a ready capability to adapt to new, different or changing requirements." In a manufacturing context, it can apply to either production growth and volume increases or physical changes to an existing product. However, machine builders, systems integrators and their customers often have different definitions and ideas about what it means to be flexible.

"The key is to define what the end user thinks of as change," says Dan Alcombright, P.E., manager of programs and project engineering at Remmele Engineering Inc. (St. Paul, MN). "Some users see flexibility as the ability to manufacture many [different products] on the same manufacturing line with little or no changeover. This ability greatly reduces cost by eliminating or reducing inventory cost.

"In some cases, the end user sees flexibility as the ability to reconfigure the equipment years down the road to support new product platforms," adds Alcombright. "Frankly, this kind of flexibility is a crapshoot since the end user can rarely define, in engineering terms, what flexibility is required.

"The greatest success we have seen is where clients measure flexibility in terms of manufacturing a family of [products] on the same manufacturing line. For example, in the window manufacturing industry, flexible cells can adjust on the fly to literally hundreds of sizes with only seconds of lost production time and no tooling changeover."

Paul Zielbauer, technical service manager for linear motion and assembly technologies at Bosch Rexroth Corp. (Buchanan, MI), also believes that there are many ways to view flexible assembly. "It all depends on the type of flexibility a manufacturer needs," he explains. "If [it] needs to make multiple products on a single line, that's flexibility to that customer. If [an engineer] needs to reconfigure his line, then that's his definition. It might also mean that he can add automatic stations in place of manual workstations already integrated in the line, or that he can quickly change over from making one product to another, with very little effort."

Pros and Cons

There are numerous pros and cons to flexible assembly that manufacturing engineers must carefully consider. For instance, traditional assembly equipment is limited by attributes such as processes, motions, precision, cycle rates, part size and weight, and component feeding. However, a dedicated machine is typically less expensive than a flexible machine.

"There are often overlooked costs associated with flexible assembly," warns Zielbauer. "For example, a lot of cost goes into the wiring and plumbing of a conveyor that will need to be redone if the line is reconfigured. There are also programming changes which could result in considerable additional cost, if the changes are significant."

And, compared with a dedicated machine, a flexible machine often has a larger ratio of size vs. throughput.

As production volume increases, flexibility decreases. Generally, if annual production volume is at least 5 million units, or if the product's market life is expected to be at least 5 years, manufacturers are better off with a dedicated machine.

"Typically, there are premiums associated with flexible equipment," notes Sortimat's Speiden. "At a minimum, nest costs [tooling that the product is assembled around] are significantly higher because of the sheer number of them required."

In addition, Speiden says manufacturers need to have "a plan to make service parts after the retool if their customer contract demands it. As the demand for quality approaches zero defects, newer technology will be required to meet this demand that was not available when the machine was initially built."

Motion control plays a key role in achieving flexibility. Instead of using pneumatic actuators or cams, servo-driven actuators enable engineers to change motions simply by entering a new value in the control software. Servomotors are more expensive than pneumatic actuators.

Efficiency is another trade-off that needs to be carefully evaluated. Highly flexible systems may be unable to match the speed of dedicated systems.

Component feeding can also detract from the inherent advantages of flexibility. Indeed, if component supply via feeder bowls, reels or trays is significantly different from one product to the next, flexibility may be too extreme to be practical.

Fixtures are also critical to flexible assembly systems. Adjustable fixtures feature one or more sides built on slides so that the length and width of the fixture can be quickly changed. Multipiece, adjustable fixtures are more expensive than off-the-shelf fixtures.

"The disadvantage is everything is done on this one pallet, making it expensive and hard to work with," says Eric Pasman, technology design section head at Sortimat. "On a dedicated machine, you can split the fixtures, making processes simpler and more cost effective. [Running] many processes on one pallet is costly, makes the stations more difficult and complicated, [and makes] pallets expensive to replace."

Myths and Misperceptions

Many myths surround flexible assembly. For instance, people often underestimate the time and money it will take to change a system from one setup to the next, simply because they didn't plan on change from the very beginning.

"Most people think it is simple and quick, but it is not an easy task," warns Paul Nordin, senior project manager at Sortimat. "Flexible assembly gets compared to buying a commodity-type machine tool. The user feels that a new fixture is all that should be needed to cut chips or perform assembly for a new product. However, cutting chips does not include all of the quality checks and performance testing required of an assembly system. These are all discretely designed for the product at hand, to save up-front costs, and need to be redesigned and implemented at the time of retool. Most [assembly] processes will require changing more than just the pallets."

"A little foresight goes a long way, and may even determine which type of [machine] someone buys to start," says Bosch Rexroth's Zielbauer. "For example, modular assembly conveyors are much easier to expand or reconfigure than rotary dial machines. An extruded aluminum frame makes a big difference compared with welded steel, and the framing elements left over after a conveyor reconfiguration often become workstation bases, guarding, machine bases or other factory structures.

"And even within a family, an end user may choose roller chain instead of an anti-static belt, because he knows it will take a little more time to weld new belts than to replace the chain," adds Zielbauer. "Or he may buy completely assembled transverse conveyor units and build a line out of them, knowing that he can easily reconfigure these later."

According to Apex Automation's Zimmerman, many engineers make the mistake of believing that they can predict manufacturing requirements 2, 3 or 5 years into the future to meet product development demand. When that happens, key subsystems, such as control systems, often become obsolete down the road.

"These systems are then returned for expensive and time-consuming upgrades, reconfiguration or replacement," says Zimmerman. "The incorrect definition of flexibility can double the cost of an automation project, greatly increase lead time and increase risk. The flexible system may have so many competing requirements that it is not able to meet any one requirement very well."

Zimmerman believes engineers should define the level of flexibility that will directly add value to the business. He recommends conducting a technical risk evaluation that assesses the costs, benefits and risks for each general approach to flexibility and verifies that the intended flexibility will lead to reduced risk and increased net present value.

Planning Is Critical

To be successful, flexible assembly requires detailed planning. However, that can create numerous headaches for engineers. "Flexible assembly is a worthy goal, but trying to achieve it can be frustrating," Zimmerman points out. "Dedicated automation allows engineers to address current or short-term issues only. They may have more issues to address, but each is relatively simple.

"Engineers, management and accounting need to remember that true flexible assembly, like zero defects, will not be realized," says Zimmerman. "The best you can hope for is that your results are asymptotic to the goal. One management team I worked with in a Fortune 200 company insisted on using the term ‘universal assembly machine.' I never was successful in making them realize that universal was the goal-multipurpose was the reality."

Companies that are successful in justifying the higher cost on the front end of a flexible project don't adhere strictly to time-based payback periods. "They can make a logical business case for the projected residual value of a system after its initial use-many times without being able to point to the bottom line, hard-money justification," notes Zimmerman.

"Even if you know ahead of time what kind of changes will be made to the product in the future, it's a lot harder to design for something flexible than for a dedicated machine," says Sortimat's Pasman. "The process is harder because you need to consider building in ‘ranges' of measurements to compensate for any hypothetical product changes.

"But, once the initial concept is figured out, you use this as a base later on to retool," adds Pasman. "The ranges will be used as design ‘cushions' when new product comes out later to make the transition smoother." Pasman says it's easy to decide what level of flexibility you need if you know-or have some idea of-the potential changes in the product. Of course, this is not an easy task when you're thinking 5 years down the road.

"While it might cause fewer headaches to buy one machine, throw it away later and buy a new one, if flexible technologies are planned correctly, with an eye toward changeover, flexibility and the speed with which that can be accomplished, then any headaches can be minimized without wasting the initial investment," claims Bosch Rexroth's Zielbauer. "The most headaches of all come when companies try to buy cheap equipment and expect it to run forever."

Going the Flex Route

Implementing flexible assembly is not as difficult as some people think. A flexible system can be started with automated stations alternating with manual stations on a workholding and transfer system. It can be further automated with the addition of more stations or modules later.

"Automation is more flexible when you begin with a semiautomatic, low-volume assembly machine," says Sortimat's Speiden.

The first stage of flexible assembly is most flexible. As volume increases and automatic stations are added, a machine becomes less flexible and requires more time and effort to retool.

During the initial product launch, a manufacturer may use an operator to load certain components, and add mechanisms to automatically load them at a later time. However, there are challenges, such as balancing the required output of the equipment with an operator's ability to load parts. Assemblers may discover that they have to rework a machine when ramping up to a higher rate.

When planning a flexible assembly system, engineers often overlook a variety of parameters and measurements, says Speiden. For example, the speed of product transfer may need to be increased. Or the pallets could be too small for the next product.

The process of changing over from one product to another often depends on how much flexibility was planned into the original system. "At one end of the spectrum, changeovers can be as simple as programming or as involved as complete system reconfiguration," says Zielbauer. "It really depends on how much flexibility is built into the system from the beginning.

"For example, if someone knows that they will be running multiple products on a single line, they may outfit workpiece pallets with product identification tags that allow the system to separate the different products onto different conveyor lanes as needed. In this case, much of the work is advance programming. Of course, this intelligence can easily be added later, as can new pallets with new ID devices.

"The other end of the spectrum is complete reconfiguration," says Zielbauer, "in which you tear down the existing system and rebuild it with the existing modules and the additional new parts or modules to get the desired system. This is not necessarily faster than buying all new equipment, but it makes use of the initial investment and is much more cost-effective.

"In the middle of this would be interchangeable workcells, which might be wheeled into place on a conveyor line as needed, then removed and replaced with a different cell when product changeovers are required," explains Zielbauer. "For a company that wants to start small, then add cycle-independent loops or automatic processes as it grows, adding to a system is quite easy and fast, as long as floor space is available."

Reusing Old Equipment

Reconfigurability is a key attribute of flexible assembly that varies from application to application. The extent to which flexible assembly equipment can be reconfigured depends upon the amount of detailed planning that went into the original system, how reasonable the expectations are, how "flexible" was defined at the beginning of the project, and the robustness that was built into the system at the start.

"It doesn't make sense to reconfigure a transfer device," explains Zimmerman. "If it was not in service beyond its useful life; if it was well-maintained; if critical components, such as electronic controls and servo systems, are still supported by the manufacturer; and if there is still a requirement for processes that fall within the equipment's capabilities, it can be reused. Many times, one or more of these criteria are not met, and the equipment needs to be retired, even though it was installed as a flexible system."

"The amount of reconfiguring and reuse that can be done depends on the design and the difference between the existing customer assembly and the new one," says ATS Automation Tooling Systems' Cybulski. "Robots are reused the most because they are easily reprogrammable and are an expensive part of an automation system. Retrofitting an existing system can cost anywhere from 50 percent to 80 percent of the cost of a new assembly system."

Sortimat's Nordin recently conducted a study of flexible automation. He discovered that 60 percent of the capital could be reused and the cost was about 50 percent to 70 percent of the original contract. "In the telecom industry, 90 percent of things can be reused for 10 percent to 20 percent of the retool costs," says Nordin. "But, these are always reused in the same product niche of the same industry. This is not a portrayal of ultimate flexibility. Greater expectations of how flexible machines can be will lead to increased costs during reuse and retool."

Once products and processes become obsolete in a flexible assembly environment, most manufacturers "pitch it, nine times out of 10," Nordin claims. "Once [in a while, they'll] chop them down, lower the rate, and make spares. It is rare to bring them back and use them after 10 years."

However, some observers claim there are no limits to reconfiguring and reusing some equipment. "Our modular framing system lets people add on anything they need or reconfigure existing lines and reuse the leftover parts," notes Bosch Rexroth's Zielbauer. "The most important consideration is to plan well enough so that you can take best advantage of the initial investment. If you do, it will always be less expensive than buying an entirely new system. Some of our customers have even found it less expensive to reconfigure or move existing systems than to outsource their manufacturing overseas."

If necessary, a conveyor or other piece of flexible assembly equipment may be disassembled into manageable sections and stored, Zielbauer points out. The structural elements, for example, could form the core of a new lean manufacturing or manual assembly cell. If a plant is closed, systems are typically auctioned off or given to another facility of the same company. "If a plant is acquired, we are often called in to redeploy a system that was already in place," concludes Zielbauer.