New approaches and processes are required to deal with increasingly complex wiring system integration challenges, including pressure from regulatory mandates and the need to do-more-with-less.
BY JOHN LOW
The aerospace industry has long been perceived as slow to adapt to new wire harness engineering technologies and processes. Many enterprises rely on systems created in-house, with each tool being structured to support a very specific design/build process. Departments and divisions responsible for defining the electrical content of vehicles and integrating the electrical systems have one of the most demanding jobs in vehicle design.
 |
While these organizations are often understaffed relative to the scope of their task, they have the distinction of being expected to manage the highest number and rate of design changes. Companies, especially OEMs, are finding it difficult to adopt new wire harness technologies and processes while maintaining the quality of their deliverables.
Until recently, electrical system and wire harness development tools have not been viewed as an adequate solution for developing complex electrical wiring systems. Now, aircraft systems and wiring designs are making a transition from traditional mechanical and pneumatic systems to electrical, and the entire electrical architecture of the vehicle is becoming more complex. Companies are finding that their existing solutions do not provide the needed engineering capability. But powerful new technologies are emerging, and many companies are taking another look.
Open to change
Integrating a multitude of systems and subsystems into a platform is a difficult, time-intensive task that involves a host of engineering decisions. Too often, the impact of a design error in a wiring harness is not discovered until the harness has been manufactured, tested, and installed. Then–after all the efforts to reduce development risk and follow rigorous, disciplined development steps–the program objectives slip and electrical system integration becomes the critical path. It is a blow few careers can withstand.
Even as electrical systems increase in complexity, emerging regulatory standards are taking aim at wiring systems. The relatively new Electrical Wiring Interconnection System (EWIS) regulation addresses the decisions made during the system integration process, as well as the subsequent maintenance and service of the resulting wire harnesses. Companies are learning that regulatory mandates provide ample reason to re-evaluate their internal development processes. In doing so, these companies are seeking more efficient and repeatable methods that add value rather than cost.
In many established electrical design flows, there are issues that complicate electrical system integration and add risk:
- System and subsystem data is typically described in diverse, heterogeneous media, such as text documents, spreadsheets, sketches, design handbook guidelines, and more. Data in these forms is difficult to recover and aggregate, and does not lend itself to “what-if” testing of alternative integration approaches.
- Engineering intellectual property (IP) for wiring integration often resides in the heads of skilled, experienced engineers. It is an ephemeral form that makes it difficult to capture and leverage the organization’s crucial best practices. In turn, it is difficult to automate decision-making for the purpose of ensuring consistent design quality.
- In the real world, design changes occur constantly. Reacting promptly and accurately to these changes is a challenge. Of course, any delay within the integration group can have a ripple effect on manufacturing planning. That puts the schedule at risk, with attendant costs.
Model-driven solutions
Around the world, companies are beginning to realize that even though wiring integration is a key part of wiring systems development, true process innovation means improving the entire design flow. For example, if system and subsystem design data is captured in a form that makes re-use easy and can drive downstream decision making, then the wiring integration process benefits. If the underlying data truly represents the actual electrical model being described, then early design verification can occur at any point in the design process, thereby addressing the design change issue. If IP is captured as design constraints, then the documented information can be used during wiring integration to synthesize (or “derive”) a physical wiring implementation.
These characteristics all support the Model-Driven Development (MDD) process, a methodology that is fast gaining acceptance in the realm of wiring system design and integration. Several commercial software sup-pliers today provide solutions to support the model-driven development process.
The MDD approach allows the designer to capture the system and subsystem design data in a form that describes its connectivity, operating characteristics, and associated design constraints, such as EMC requirements. This logical design data “package” forms an electrical model that can be used to study behavior at a very early stage. By analyzing the model in various pairings with other models, it is possible to understand the combined effects before wiring integration starts. The modeling environment, for example, easily spots unwanted electrical interactions. Design issues can be identified and corrected at an early stage, averting disaster later when hardware prototypes have been built at great cost.
As the design process advances into wiring integration, the logical system/subsystem data and the appropriate wiring integration guidelines (including regulatory and design constraint IP) combine with geometric constraints to synthesize the wiring model. This model is a complete platform-level wiring architecture that reflects optimized signal routing, correctly sized wires, and proper grounding implementations.
 FIGURE 1. The flow of a commercially-available toolset currently in use in many design departments. |
Wiring synthesis is not simply wire routing–it is a process that automatically implements design intent based on a deterministic, constraint-driven process. A fundamental precept in the model-driven development process is that, at any point, the data is in a form that enables quick and efficient electrical verification (also known as electrical virtual prototyping).
At the beginning of the process (the orange box in Figure 1), electrical systems and subsystems are captured in a form that pictorially represents signal connectivity. Parameters that describe the operating characteristics of the equipment and signals are added to this connectivity; attributes, such as power and signal characteristics, EMC classifications, and other details, become part of the high-level description. At this point, the result is a fully defined logical signal model. The model can be analyzed for FMEA, sneak circuits, and a multitude of other effects. Because there is no physical or wiring information as yet, necessary changes can be made quickly and cheaply.
Proceeding to the next step in the design cycle, the designer uses geometric and environmental constraints–often extracted from the 3D MCAD domain–to create a topological harness model. Some areas or zones may have special requirements for the wiring system: high heat, rotor burst areas, etc. This information is described with easy-to-read constraints that the engineer applies to the selected locations within the topological representation. These constraints are reusable and can be put in a library for future use on other programs. Furthermore, these constraints can be grouped into a rule set. For example, a rule set for “flight controls” would include all the constraints applicable for flight control systems.
Constraints are a core concept in the model-driven development process. They drive part selection, determine correct signal routing based on EMC compatibility, define splicing, and implement grounding strategies and other system integration design tasks. From an EWIS (external wireless instrumentation system) perspective, constraints automate decision-making in a way that guarantees adherence to design mandates and regulations.
When the foregoing steps are completed, it is time to synthesize the integrated wiring model, as shown in the red box in Figure 1. Wiring synthesis is a process of evaluating the system and subsystem information, considering geometric, environmental, and electrical design constraints and then automatically implementing the wiring model.
Figure 2 is an example of the wiring synthesis output. The output from wiring system integration is wiring diagrams. These images depict the physical “as-built” engineering configuration of the wiring information. In a model-driven development process, this wiring information has already been determined via constraints and other design decisions previously described. But the wiring diagram itself is a human-readable visual representation of the wiring integration results.
In effect, the modeling process automatically documents its own results.
Rather than a manually created document that is susceptible to data re-entry errors, the wiring diagram is a generated artifact of the wiring model. If the model data changes, it is a simple matter to regenerate the document. By definition, the freshly regenerated wiring diagram will reflect the current revisions in the wiring data.
Virtual prototyping, analysis
At any time during the development process, designers can elect to perform electrical virtual prototyping and analysis. The electrical data is always available in a form that makes analysis simple, and logical and physical steady-state or time-domain simulation can be undertaken even as the actual design evolves. The design team can easily detect problems early in the process, and then make changes while they are still relatively inexpensive to implement.
 FIGURE 2. The wiring synthesis process maps the details of the aircraft wiring model. Here, the three views depict the nose undercarriage system at various levels. |
The last step in the wiring design process is manufacturing planning, depicted in the green box in Figure 1. This step allows the planner to use the synthesized wiring architecture and the harness topology, and to embellish the data to create a manufacturable harness with all its supporting documentation. Since wiring information, such as gauge, color, and insulation have been determined in the previous integration step, this data can be leveraged to automatically select contacts, plugs, seals and other parts that comprise the harness manufacturing bill of materials (MBOM). In addition, artifacts (such as formboards and data to drive the downstream manufacturing processes) can be easily created and shared.
As mentioned, the goal is to identify design issues as early as possible; however, there will inevitably be design changes of various sorts that occur along the entire design cycle. Leading commercially-available tool providers understand the critical importance of a data-centric architecture as the foundation for design tools. This ensures that changes and updates can be propagated readily, at the lowest possible cost.
Meeting integration challenges
Many factors have added up to create a tipping point in the aerospace industry’s willingness to consider commercial vendor-supplied wiring design tools. Companies are investigating new approaches and processes to deal with increasingly complex wiring system integration challenges. These challenges include pressure from regulatory mandates and, equally important, the need to do more work with fewer people.
Model-driven development has been standard practice in the IC and PCB worlds for years. The past few years have seen this methodology also emerge in the wiring systems development field, because today’s wiring integration challenges now approach the complexity of some IC chips and circuit boards. Enterprises using this modern philosophy are now reaping the benefits of model-driven development, including:
- Reduced engineering/ECO costs;
- Engineering data existing in a form that can be analyzed and verified at any step of the process;
- Increased time and opportunity to explore design alternatives;
- Increased productivity thanks to reusable proprietary IP;
- New engineers “come up to speed” more quickly, and rigorous design rules and constraints help them avoid costly errors;
- Model-driven development is a process–ready for use today–that brings true innovation to the wiring system integration process. CS
JOHN LOW is aerospace business development manager at Mentor Graphics Corp. (www.mentor.com). He started with design and manufacturing engineering responsibilities on rocket propulsion systems, followed by work with The Boeing Company in a variety of design engineering roles. For the last 15 years, Low has been with the Integrated Electrical Systems Division of Mentor Graphics Corp.
