- SPECIAL REPORTS
Prior to the downturn in the telecommunications industry, numerous barriers to automating optoelectronic device assembly existed. Such barriers include unusual form factors, visual and physical barriers within packages, lack of references for machine vision, low volumes, short product life cycles and nonrepeatable labor-intensive processes. And during this period, full automation will still be less viable. So package volumes will continue to favor manual or semiautomated assembly methods.
In any manufacturing environment, certain fundamental principles are used to determine the type and class of assembly to
be performed. These principles are based on technology, complexity, density and volume. Volume is the key criteria.
Optoelectronic assembly covers the full spectrum of volume categories-from low-volume, high-mix through high-volume, low-mix production.
Typical assembly volumes for products containing optoelectronics are in the 50 to 1,000 units per year range. This translates to approximately 1 to 20 units per week. Even high-volume optoelectronic applications only require assembling approximately 5,000 units per year. Much of the current optoelectronic assembly capacity is dedicated to low-volume, high-mix applications.
A Matter of Class
There are three classes of assembly for manufacturing environments. Class A is fully automated assembly. There is minimal human interaction with the product being assembled. Class B is semiautomated assembly. Here human interaction is required to transport the product from workcell to workcell and to load the work-in-process (WIP) into the workcell for processing. Class C is manual assembly. It is well suited to a low-volume environment by providing a high level of flexibility and asset utilization, even with product revision.
Manual assembly has its drawbacks, specifically in terms of WIP and yield. In the current environment, WIP can be a positive and a negative factor. WIP is a positive where minor product engineering and process changes are common, because the changes can be implemented real-time without significant tooling modification and a large number of rejects. WIP is a negative where major engineering and process changes are common.
The typical yields for a Class C process are less than 50 percent.I However, effective workcell design, selection and layout, along with the right process, tools and assembly practices, can positively impact the production yield of a Class C process.
Class B production offers a higher level of automation, which is well suited to a low-to medium-volume manufacturing environment. One automated Class B workstation can typically replace four manual Class C workstations. The workstation yield can increase by as much as 30 percent.II However, due to decreased flexibility and increased tooling, Class B assembly is only suited to situations where the product design is fixed and stable. Some level of product and production standardization must exist to minimize in-process changes. Though there is the benefit of reduced WIP through increased throughput, rejects can also be produced faster. There is also a significant increase in capital and specialty tooling costs for upgrading to a Class B process. The typical increase in capital costs for a Class B process are approximately 5 to 10 times that of a Class C process.
Effective design of an optoelectronics assembly process requires understanding the assembly requirements. The first step is to define the work to be accomplished. This includes a 5-year forecast of the number of units to be built by product, the component types and quantities associated with each product, and the kinds of substrates and packaging requirements necessary for each product. The second step is to develop the manufacturing requirements. This includes defining the components per the bill of materials (BOM), defining the substrate size and material, defining the number of different unique values within each family of component, analyzing the components by technology, and developing an initial production schedule that defines the production requirements and product mix per month. The third step is to develop a product characterization based on the BOM and estimated production forecast for the third year of production. From the third-year product characterization, a full-capacity analysis can be performed to determine the class process.
Once the class process has been determined, equipment and tool requirements should be decided. Equipment and tool requirements are based on several factors, including product complexity, product handling requirements, design structure and stability, desired quality and reliability.
Workstation and Facility Design
The assembly workstation is the foundation for assembling optoelectronic components. The workstation must support the assembly process and the task sequence. Because these assembly tasks are largely performed or heavily influenced by humans, sound ergonomics must be employed to minimize fatigue. Fatigue contributes to errors. Conversely, a well-designed workstation will help technicians maintain focus and fine motor control throughout the shift, which positively contributes to product quality.
Workstations should be space-efficient and adaptable to future needs. Correctly designed workstations are a production tool that can enhance the production process by bringing order to the assembly stage. Workstations must also position, secure and protect sensitive equipment while preventing damage to components by static discharge or contributing to assembly error by introducing unwanted vibrations or oscillations.
When designing workstations for the production process, a number of basic factors should be considered. First, define the process flow, and how assembly will be accomplished. Will it be a build-in-place process, where one technician at one workstation will complete all process steps? Or will it be some form of progressive build? Where and how will units be tested? Will troubleshooting and rework be performed at the same station, or at different ones? Does management follow lean manufacturing, and if so, what impact does that have on the design of the layout, and the movement and storage of inventory?
Determining your process flow will largely be a function of several issues: product volume and mix; number and complexity of process steps; labor costs; and sensitivity to movement
Depending on the product, a build-in-place approach may be specified. With higher volumes and improvements in design, processes and tools, a progressive build may be desirable, particularly if certain process steps can be accomplished using lower cost labor than other process steps.
In progressive assembly environments, conveyance between workstations also requires study. Modular workstations usually allow standalone assembly islands. Or they can be linked to create progressive assembly lines. Conveyors, ball transfer systems, pallet and magazine transfer systems, and flow racks can be incorporated into workstation systems to help move the product.
Integrating Tools and Equipment
Various tools, equipment and materials may be required to build a particular product. While many of the tools and equipment may be acquired off-the-shelf, the optoelectronics industry still uses a lot of proprietary equipment. A lack of package and process standardization has driven this.III As Class C processes evolve to Class B processes, the type and amount of semiautomated equipment used will increase. Specialty semiautomated equipment may be bench-mounted, while others may be supported on their own frames or stands. Operators' work areas should be uncluttered, with only the needed items contained within the work envelope.
Equipment can be organized to maximize space efficiency. Exploiting vertical space at the workstation is the best way to reduce the overall size of the workstation while keeping essential items in reach of the operator and addressing ergonomic concerns.
To best use workstation space, it is often helpful to remotely locate components of the system that do not require operator interface. This frees up space for other process equipment, tools or materials.
Equipment can be positioned anywhere in the X, Y or Z axes. There are five options for placement of equipment in the Z axis:
- On the work surface.
- Inset or recessed in the work surface.
- Suspended below the work surface.
- Above the work surface, shelf or arm.
- On the floor or on a roller shelf.
Depending upon the sensitivity of the tools or equipment used for assembly, and the assembly tolerances for the device being assembled, vibration at the work surface could negatively impact yields. Vibration can be introduced at the work surface in various ways:
- Transferred from the floor.
- Transferred from motorized or pneumatic tools.
- Induced by the technician.
- Induced by adjacent workstations.
- Workbenches with independent vertical space management offer the highest resistance to vibration isolation.
- Standard workbenches with integrated vertical space management offer moderate resistance to vibration.
- Modular cantilever workbenches offer the least resistance to vibration, but maximum flexibility of configuration and adjustability.
Workstation design will be affected, depending on how demanding the need is for a vibration-free work surface. Several techniques can be used.
When specifying workstations, consider what kind of reconfigurations may be required over the life of the workstation or the project. Product changeovers; new product introductions; changing assembly, test or packaging processes; the need to incorporate new tools or equipment; and the impact of multishift operations or staff changes may drive reconfiguration needs.
Typical reconfigurations include adding, removing or relocating shelves, tool and equipment holders, lighting, utilities, computer peripherals and other accessories. Also, converting from single-sided to double-sided units or vice versa, and converting from standalone to in-line workstations are common.
Consider these factors regarding reconfiguration issues:
The Need For-What are the chances that changes in product design, build processes or use will change? How far out does your planning horizon extend?
The Ability To-What are the actual mechanics and costs of reconfiguring the workstation? Can it be converted from single sided to double sided, or vice versa? Can units be linked to create multistation in-line configurations? What additional parts are required? What parts already purchased will no longer be needed?
The Ease Of-What tools are required? How heavy are workstation components? How many people does it take to reconfigure a workstation? What ergonomic or safety issues might be encountered? What are the costs involved in components, labor and lost production?
Ergonomics should to be considered throughout the entire product use cycle, which includes workstation installation, workstation use and workstation re-use.IV
Fast and easy assemblies, the ability to customize during use, and rapid disassembly, relocation or reconfiguration are all areas of ergonomic impact. However, ergonomics will most positively impact cycle times and yields.
Key areas for ergonomic focus are the reach zone, establishing the correct work height, and setting and changing work surface heights relative to the work being performed. As processes evolve, task definitions change or new workers are introduced, these three issues need to be addressed.
The reach zone is subdivided into three areas in the horizontal plane: the optimum work area, the optimum grab area and the maximum grab area. The optimum work area is a 10- by 10-inch box located 5 inches in front of the operator.V
Processes and task sequences need to be contained within the envelope defined by these areas to create an efficient workcell and minimize fatigue and strain on the assembler. Tools and equipment should be organized within the reach zone according to frequency of use. All tasks should be accomplished within the optimum work area if possible.
It may seem most desirable to have a single operator or assembler perform all tasks in the assembly process. But the degree to which the tasks can be contained within the optimum work area will affect the ability of the operator to avoid muscle fatigue. This affects productivity, accuracy and yields. It may be desirable to divide the assembly process over more than one assembler, so yields at each workstation can be optimized.
Height of the workpiece, control panel or other device the operator must routinely interface with affects the operator's posture. This affects the potential for fatigue, strain or repetitive motion injuries. The workpiece height refers to the Z height on the workpiece where the specific work is being performed. This should not be confused with work surface height. Work surface height is the plane on which the device, or its fixture or holding device, is placed. The work height may be above or below this point.
Work height is the relationship to 2 inches below elbow height. For light assembly tasks, this is increased to 2 to 4 inches. Heavier assembly is 4 to 6 inches. In cases where work height may need to change for a single operator, an adjustable-height workstation or some type of fixturing that provides variable elevation may be used.
Major differences that affect work surface height are based on whether the assembler plans to sit or stand. It's possible to set up a workstation for a single person that has split work surfaces. This allows for side-by-side work if some tasks are better performed standing vs. sitting. For men performing standing work, ideal work surface height ranges from 29 inches to 43 inches, depending on the type of work being performed. For women, the range is 27 inches to 41 inches.
Workstation design should provide for work surface height adjustments that will accommodate the broadest range of individuals based on their particular anthropometric characteristics.
Cost Of Ownership and Use
Cost is the most significant nontechnical factor in determining the class and type of equipment to be implemented. Traditional capital costs, direct labor costs, return on investment, cost of quality, increased capacity costs, cost of ownership and cost of use should be investigated. Projections should be based on the 5-year forecasts for each product in conjunction with the inherent capability of the intended process.
Cost of workstation ownership includes purchase price, cost of use and useful life. The cost most easily focused on is purchase price, but this is typically the smallest contributor to the cost of ownership. Because workstation products that meet the criteria for adaptability typically have a long useful life, costs of obsolescence are usually of little concern except when a great deal of process-specific customization has been specified. Even then, the basic architecture or framework of the workstation can typically be used for many years. Cost of use then becomes the primary cost factor. Contributors to the cost of use are: cost of installation, cost of daily use, cost of re-use and cost of space.
I. Weiss, Stephanie A., "Photonics and Volume Manufacturing: The Automation Crisis." Photonics Spectra magazine, 2001, 6, 35, pp. 98-110.
II. Pate, Bryan, "Automating Optoelectronic Packaging." Journal of Surface Mount Technology, 2001, 7, 14, pp. 7-12.
III. Weiss, Stephanie A., "Photonics and Volume Manufacturing: The Automation Crisis." Photonics Spectra magazine, 2001, 6, 35, pp. 98-110.
IV. Dr. Alan Hedge, "Ergonomic Benefits of Workstations in the Product Use Cycle." Internal Memorandum, March 2001.
V. The Canadian Standards Association CAN/CSA-Z412M89, "Office Ergonomics: A National Standard of Canada," Oct. 2001, pp. 40-46.