Product integrity occurs when performance, schedule, and affordability converge throughout the product life cycle. The first critical stage in realizing product integrity happens early in the product life cycle during design and development; a second and no less critical stage occurs later, during the transition from development to production. Early in the process, the relationship between design intent and process capability must be established and understood. As the design matures and transitions to production, it must be manufactured in a repeatable and affordable way by an extended supply chain. Achieving these seemingly intuitive objectives continues to be elusive for much of the aerospace and defense industry.
The industry has invested extensively in lean and Six Sigma. Best practices, including “tollgate” systems, supply-chain collaboration, and model-based design, have been broadly applied. Yet major programs continue to exceed cost and experience delays. The U.S. Government Accountability Office’s (GAO) March 2008 report, “Assessments of Selected Weapon Programs,” identifies extensive cost overruns and delays for numerous major weapon systems--just as the past five such reports have. The civil aviation side is not immune from similar problems, as recent well-publicized issues for major U.S. and European programs illustrate.
Analysis of these problems raises questions about current approaches to best practices and how they might be improved. One impediment is the difficulty in capturing fundamental knowledge of design practices or constraints, as well as process capability, at the component part level. Without a clear grasp of such knowledge, it’s virtually impossible to apply that knowledge early enough in the product life cycle so that downstream problems are avoided during production.
Another potential shortfall is the level of rigor and discipline applied to detailed product understanding throughout the product life cycle. The concept of characteristic- level accountability is an important aspect of current aerospace industry quality standards, including AS9100 (quality management systems), AS9102 (first-article inspection), and AS9103 (variation management of key characteristics). Finally, there’s the need to effectively flow essential design and process requirements through multiple tiers of a supply chain that is growing in both size and level of responsibility.
It’s been observed that characteristic understanding is to product definition as atoms are to physics. A product is the sum of the dimensions, geometries, notes, and specification references that define what it is, how it will perform, how it must be made, and how it will fit with the next-higher assembly. Depending on the product, characteristics might go beyond physical attributes to include such things as input/output relationships. Characteristics are discretely measurable and comprise the building blocks that create product features. If a hole is a feature, the diameter, depth, and surface finish of the hole are its characteristics. Individual part characteristics are the center point where design intent and process capability come together. They form the bridge between abstract design and tangible hardware.
As the illustration in figure 1, below, shows, a discrete part’s characteristics are determined by its relationship to the larger product, which is the result of the overall system requirements. Converting these part characteristics into a real part is a factor of how capable the selected processes are in efficiently making this conversion. Starting with the part and its characteristics, it’s possible to look upstream at the design/development process and downstream at the operations/supply chain process. The upstream process is one of knowledge capture and application; the downstream process is one of knowledge execution. Both are dependent on a clear understanding of characteristics and their relationships to design requirements and process capability, coupled with a disciplined approach to making those relationships work. Each has a set of emerging approaches that, when applied, will substantially improve product integrity.
A quick web search reveals that virtually every major aerospace company is pursuing design for Six Sigma (DFSS) or one of its variants. Design for manufacturing and assembly (DFMA) and the model-based enterprise (MBE) are two prominent approaches to achieving integration between design and manufacturing in pursuit of the DFSS vision.
For the model-based enterprise, a consortium that includes BAE Systems, Boeing, GE Aviation, Lockheed Martin, Raytheon, Rockwell Collins, and Rolls Royce offered an overview of its activities at the 2007 Defense Manufacturing Conference. The consortium’s ambitious efforts seek to leverage the MBE concept to achieve “integrated, agile, and flexible engineering, manufacturing, and sustainment… utilizing common processes and integrated systems/tools/data across the extended enterprise.”
In examining the examples provided at the conference, it appears that if MBE is geared to enabling DFMA, there’s considerably more emphasis on the “Assembly” than the “Manufacturing.” The highlighted MBE successes focus on what appear to be post-design activities at the part level, emphasizing how components come together for higher assemblies, ergonomic factors, and interference avoidance. These are very important considerations, but they don’t explain the creation of a model that yields a producible design at the part level.
Part-level producibility is the key to product integrity. An assembly might come together seamlessly through the use of models, but the individual components that make up the assembly drive performance, schedule, and affordability. As companies have begun to recognize this, they realize that achieving their DFSS or DFMA objectives at the part level depends on three factors:
• Knowledge capture and reuse. Understanding process capabilities and limitations, and documenting known design practices and constraints, are essential elements that are seldom well-documented. This captured knowledge, when coupled with a common interpretation, consistent application, and ease of access and/or reuse, provides a common starting point during design.
• Knowledge application. Applying and integrating knowledge so that characteristic-level process capability can be weighed against design requirements is a critical step in the realization of DFSS.
• Knowledge exchange. The ability to take the knowledge, apply it in such a way that informed decisions can be made, and adjust accordingly is a function of collaborative exchange. However, to be effective, exchange among the disciplines involved in product development must be timely and integral to the process.
Making all this happen at the part level requires discipline and commitment to the process, but it’s sometimes missing even in the model-based enterprise. Bringing manufacturing engineers into the design process is a popular approach. However, it’s where and when they enter the process that determines their effectiveness. The historical tendency to “throw the design over the wall” has clearly lost favor. The more recent--and more enlightened--approach provides for a producibility assessment (often near the end of the design process) as a bolt-on activity during an obligatory tollgate review. This method has limitations because the manufacturing input often means changes to the design, which in effect creates “rework” for the design engineer. The bolt-on approach, although better than no manufacturing input, doesn’t represent a truly integral and collaborative exchange process. Making the manufacturing engineer integral to the process is clearly the best approach, but it’s also challenging--both culturally and technically. The illustration in figure 2, below, highlights the three approaches.
Overcoming the cultural hurdle of making manufacturing an integral partner to design is only the first step. Even if the process allows for the manufacturing engineer to be at the design engineer’s side from the beginning, resource limitations make this impractical. However, by merging the best of the MBE’s technical promise with an enlightened design process, the three factors of knowledge capture and reuse, application, and exchange can come together seamlessly at the component part level. The enabling technology for this merger is beginning to emerge as new software products become available.
Such a software product can serve as the technical platform for integrating process knowledge and design practices, and can also provide component part characteristic- level DFMA--with an emphasis on the “M”--in an model-based enterprise. Ideally, such a system is “CAD neutral,” allowing the manufacturing engineer to receive, evaluate, and annotate a solid model image during that dynamic period when a part approaches detailed design, but before it’s too late to make meaningful suggestions. Using the combined design and manufacturing knowledge, the system can both store and dispense essential process insights, specification references that apply to individual processes, and design constraints. It then enables the application of this knowledge as direct characteristic-level annotations to the model. These annotations are then systematically captured in a “bill of characteristics” that allows them to be managed going forward throughout the product life cycle.
One such system, known as DISCUS Design, has been applied in some pilot applications. During these initial applications, it demonstrated increased engineering productivity, shorter cycles, and designs that are ready for transition and affordable production. By bringing together best practices in both the technological and cultural aspects of product development, such a system offers a practical approach to DFSS, with results that support the goals of product integrity.
The transition to operations, which includes full supply-chain engagement, is naturally made simpler if a practical approach to DFSS precedes the transition. But even for legacy parts, designed using the “historical” process, applying characteristic accountability and verification (CAV) will have a positive effect on product integrity.
CAV is the process by which all accountable characteristics are systematically identified, documented, and verified within a given technical data package. It is the execution of the knowledge exchanged and applied earlier in the product life cycle. CAV further provides for downstream assessment of selected characteristics in support of statistical process control. The two aerospace industry standards that most influence CAV execution are AS9102 and AS9103.
A number of leading aerospace companies have adopted CAV and applied integrated systems that enable its use throughout their supply chains. GE Aviation, Lockheed Martin, Honeywell Aerospace, Hawker Beechcraft, and Cessna are among those pursuing the CAV methodology. With the growing importance of the extended supply chain to the industry, the systematic flow and interpretation of characteristic-level data are key metrics for a CAV system. The bill of characteristics--the systematic tabular listing of all accountable characteristics--is considered a best-practice enabler for CAV execution. With the emphasis that CAV places on the technical data package, a combined industry/government team, sponsored by the Aerospace Industries Association, has established a National Aerospace Standard for technical data packages. The technical data package is currently in final review, and is expected to be published this year. Among the features of the National Aerospace Standard is a set of electronic scorecards that grade a technical data package, scoring the data on such factors as identification of accountable characteristics, clear delineation of verification techniques, and joint identification of risk areas by original equipment manufacturers and their suppliers.
Characteristic accountability and verification application has a direct effect on product integrity. Through CAV, major aerospace companies have reduced quality escapes by more than 50 percent, improved accuracy and reduced cycle times for first-article inspections, and substantially enhanced the completeness of their technical data packages. CAV is enabled by a variety of tools that further enhance the flow of technical requirements and improve user productivity. Perhaps the most widely used is DISCUS Basic, which is a companion to DISCUS Design.
There’s no shortage of exciting concepts for product integrity in aerospace and defense. The model-based enterprise is clearly the direction of the future. However, as the GAO’s March 2008 report and recent civil aviation experiences have shown, there’s still a long way to go; anyone looking for a silver bullet is in the wrong business. The fundamentals of characteristic accountability--starting early and continuing over the full product life cycle--are where the journey begins and ends. As the successful application of emerging software tools and methodologies such as CAV indicate, the disciplined merger of knowledge and attention to characteristic-level details remains the essential element in realizing product integrity.