Automating Wire Harness Design In Todays Automotive, Aerospace and Marine Industry

Nick Smith
Product Marketing Director, IESD, Mentor Graphics Newbury, Berkshire RG14 2QB, UK +44-778-756-6627 nick_smith@mentor.com

Abstract
The integration of system electronics and design of interconnect wiring and is a complex task in modern transportation platforms. Competitive pressures continue to drive prices down while design complexity spirals and demand for improved quality never ceases. A critical question is "how can we accommodate increased electrical / electronic functionality and more optional choice at a competitive cost and quality?" To satisfy these needs OEMs and suppliers seek innovative solutions for dramatic productivity shifts. This paper describes the contribution Electrical Computer Aided Design (ECAD) software tools can make to the challenge. Such tools can help the design engineer develop better, higher quality designs, and indeed automate significant design tasks. After describing the design problem in more detail, the paper summarizes the history of ECAD software before explaining some modern ECAD software technologies and the benefits they can deliver. The paper concludes with guidelines designed to help ECAD tool selection.

implementation compared with traditional mechanical or hydraulic implementation. Weight and space reduction: many transportation platforms, for example military aircraft, are highly sensitive to weight and space (physical packaging). Again, electronic function implementation can have a critical impact in this area.

Adding to the problem of signal number explosion, continual market sub-segmentation causes many transportation platforms to be available in a huge variety of configurations [2]. As an example, a passenger automobile may be available with right or left hand drive, gasoline or diesel engines, automatic or manual transmission, automatic or manual climate control, adaptive or passive cruise control, and any number of other options that impact the electrical design. Because it is not cost effective and often not even physically possible to offer a single electrical design that supports the superset of all configurations, a multitude of related electrical designs must be developed, managed and built. Worse, constant design change is a fact of life in transportation platform electrical design, placing a heavy additional burden on engineering teams. Additionally, almost all industries face continual pressure to reduce costs, improve product reliability and shorten time-to-market. Transportation platform electronics / electrical systems are comparatively high cost items within the overall bill-of-materials [3], and electronic / electrical failures have become a significant contributor to escalating warranty costs [4]. The design complexity outlined above acts against reducing time-to-market. Taken together, we can see why increased electronic content, configuration complexity and design change plus the commonly known business drivers are causing the discontinuity in the nature of transportation electrical systems. Of course, multiple technical advances are emerging to address the various issues. These include: Wireless systems: electrical system complexity can be reduced by employing wireless systems such as Bluetooth mobile phone headset communications. But wireless systems currently have limited application due to reliability issues (electromagetic interference) and their inability to transmit power. Configurable components: in an effort to mitigate configuration complexity it is possible to employ electrical components such as junction boxes whose internal connectivity can be individually altered. Such components may be mechanically adjustable (for example by removing a busbar) or electronically configurable. Flat cable technology: military / aerospace platforms such as missiles have long employed flat cable technology to

Keywords: Software; design; automation; electrical; wiring;

harness; automotive; aerospace; marine.

1. Introduction
Transportation platforms such as automobiles, off-road vehicles, aircraft, rail trains and marine craft have contained electrical systems of various types (lighting, radio, heating etc) for decades. However, it can truly be said that the last few years have seen a discontinuity in the nature of these electrical systems. Rapid growth in on-board electronic content and embedded software is putting huge demands on electrical design complexity. Put simply, the number of signals flowing around each platform is rising rapidly [1]. The factors causes this explosive proliferation show no signs of abating. The main ones are: Consumer demand: personal electronics has taught consumers to demand ever-growing feature content such as navigation systems, personal entertainment connectivity and active noise reduction. Many such features can only be delivered via modern electronics. Legislative imperative: governments increasingly place mandatory demands on OEMs to conform to tough standards in areas such as safety, environmental protection and product liability. Again, modern electronics is a key element in meeting these demands. Cost reduction and reliability improvement: many features, for example power steering, can be accomplished at lower cost and with higher reliability via electrical / electronic

International Wire & Cable Symposium

266

Proceedings of the 56th IWCS

minimize physical space consumption. Flat cables are slowly becoming adopted for other applications, such as automobile roof harnesses, but tend to suffer from a lack of configurability. Digital databuses: digital databuses have significant potential to reduce wiring complexity by allowing many signals (messages) to be carried via minimal wiring, for example a twisted pair. Most modern transportation platforms now contain at least one databus network, and sophisticated software is available to model such networks to ensure efficiency and reliability.

Indeed, these may be the best choice where the electrical design complexity is very low: a domestic refridgerator, for example. Level 0 tools may be reasonable solutions for producing electrical design drawings (often called schematics), but by their nature they contain no design intelligence and produce no design data. For example, it is not possible to apply Ohms law to a line on a drawing in order to calculate a correct conductor cross section, or to use a drawing package to select a wire terminal guaranteed to be compatible with a particular connector. Such activities had to be accomplished off-line using other methods. To solve this problem the first specialized ECAD tools emerged about 15 years ago, referred to in this paper as Level 1. Level 1 tools not only provided graphical schematic authoring capabilities, but the schematic representations linked directly to engineering design data via an explicit electrical object model. At Level 1 a line on a drawing represented a true electrical connection between pins on components. The electrical connection could be associated with attributes such as length and conductor cross section, allowing true electrical design engineering to be accomplished via the ECAD software. Level 1 applications typically contain or link to component and symbol libraries, and can even swap data (example: wire length) with other domains such as 3D Mechanical Computer Aided Design (MCAD). Level 1 represents a major step forward on Level 0 for electrical design problems of even moderate complexity. Most Level 1 ECAD tools can be traced back to a related discipline: printed circuit board (PCB) design. Building ECAD tools using PCB design tools as a foundation certainly allowed valuable, productive ECAD tools to be developed rapidly. But this strategy also imposed a limitation. Many concepts needed for vehicle electrical design do not appear in PCBs. Examples of such concepts include multicore cables, clips and grommets, and configurationdependent geometries. For this reason Level 1 typically deal with a rather small portion of the overall electrical design process: they are often known as point tools. To build a complete design tool chain therefore requires the integration of multiple point tools, each fulfilling its own specific task. As can be imagined the IT task of integrating (interfacing) multiple point tools, sourced from different (sometimes competing) vendors, released on different cycles and employing dissimilar data structures, is particularly challenging. Level 1 solutions, which are common today in many large scale organizations, thus typically carry high IT maintenance costs and tend to perpetuate historical, even obsolete, design processes. This situation can be somewhat improved by moving to Level 2 ECAD tools. These tools seek to reduce integration problems by inherently covering a larger portion of the design process. They may inherently support electrical design, simulation, harness engineering and manufacturing (for example the production of formboards) as well as providing component and symbol library capabilities within a single integrated design environment. Modern Level 2 packages as defined in this paper employ file-based design data storage mechanisms, and hence have a rather small IT footprint. Level 2 ECAD applications can certainly prove very cost effective for organizations whose electrical design task is of moderate complexity: makers of off-road vehicles, marine leisure craft or gliders, for example. It can fairly be said that the development of Level 2 tools is an organizational challenge for the ECAD tool vendors. Although Level 2 may not represent a particularly large software technology leap over Level 1, few commercial software vendors have

In this paper, however, we focus on the electrical system / wire harness design process itself. Specifically the role of design automation is considered, ie the set of software applications collectively known as Electrical Computer Aided Design (ECAD) tools.

2. Five Levels of ECAD tool

Although transportation platforms have contained electrical systems for decades it has only been possible to apply computing to the problem of electrical design for about 30 years. Before that design drafting has accomplished using pencil and paper. Indeed, the electrical design process remains largely graphics-based today, partly because of this embedded psychological legacy. It is possible to identify five levels of ECAD tool, emerging in a rough chronological order. These five levels are represented in Figure 1.

Figure 1. Summary of ECAD tool Levels 0 4 Initially non-specialized drawing packages were used to document electrical designs directly from the engineers brain, essentially providing a more modern version of the pencil and paper. The drawing packages certainly improved design productivity, allowing engineers to document and change their designs comparatively quickly and neatly. The tools were also comparatively cheap to purchase and inherently flexible: drawing packages could be used to design a company logo as easily as a wiring system. In this paper such applications are referred to as Level 0 tools. It is still to find many organizations today using non-specialized Level 0 tools.

International Wire & Cable Symposium

267

Proceedings of the 56th IWCS

sufficiently broad domain familiarity to develop credible solutions spanning the broader design process. Level 3 ECAD tools provide a further step forward by incorporating newer, more powerful software technology. Chief among these is the concept of data centricity. In a data-centric toolset all relevant data, from user privileges to device connectivity to component relationships, is stored within a relational database rather than in flat files. This form of data management is central to solving some of the problems of modern electrical design complexity, particularly in the areas of configuration complexity and design change management. For example, relational database storage naturally supports functionality such as where used? queries, conditional replacement (example: replace this component if that is true) and design version comparison. Much of the advanced automation functionality described below can only be effectively delivered via a data centric ECAD tool architecture. And because data centricity is a major software architecture change Level 3 tools tend to be designed from first principles to support electrical design, hence avoiding some of the software workarounds that can damage supportability of Level 1 or Level 2 tools. However, modern software technology delivers more to Level 3 ECAD tools than data centricity. Two important examples are: Web-based integration and extensibility: although like Level 2 tools Level 3 ECAD applications seek to support the extended electrical and wire harness design flow it is inevitable that they must integrate with other domains such as 3D MCAD, Product Data Management (PDM systems) or workflow tools. Such integrations have typically been fragile from an IT standpoint, for the same reasons that integration of Level 1 point tools within the electrical design process itself are fragile. Furthermore many organizations want to extend the functionality of standard ECAD tools to support their own specialized requirements, such as custom design rule checks. It is now possible to employ web-based integration technologies such as a Service Oriented Architecture (SOA) to provide much more robust integration and extension [5]. SOA technology provides loose application coupling via stable interface contracts, while providing the impression of totally seamless integration to the user. Significant IT maintenance cost reductions accrue from such technology. Computing environment independence: new Level 3 ECAD tools will typically be written in a language such as Java which provides both computer hardware and operating system independence. The software can be installed on computers running (for example) Windows, UNIX or Linux, thus providing flexibility to support the organizations IT environment policy. This is rarely the case for older Level 1 or Level 2 ECAD tools, which are usually restricted to a single operating system. Application serving technology may also be supported, further reducing the IT burden for geographically dispersed organizations.

Introduction of this paper. In essence, they help address the discontinuity identified. Example Level 4 functionality will be described in the next section.

3. Enhanced Design Automation: Level 4 ECAD Tools

Design automation software exists to assist engineers with their creative development tasks, either by automating tasks that can be accomplished by computer rather than manually or by providing a decision support environment when the design task is beyond todays mechanization technology. This section gives examples of capabilities in both categories.

3.1 Generative design

Level 4 ECAD tools break new ground in that they provide a major increase in the degree of automation available for electrical system and wire harness design. Perhaps the most spectacular new automation capability has become known as generative design because wiring designs are automatically generated from higher level inputs, rather than interactively constructed by an engineer working with a Level 0 3 tool. We can break the typical electrical design process into three parts, as shown in Figure 2.

Figure 2. Simplified electrical design process The three basic steps are: 1. System design: construct the pin-to-pin signal connectivity for each system that will appear on the platform (lighting system, navigation system, communication system etc). System integration: implement the systems together into the physical host, transforming signals into physical wires. Harness engineering: complete the detailed wire harness design (balance splices, add mechanical components, optimize for manufacture, etc).

2. 3.

Finally we can recognize a Level 4 category of ECAD tools. Level 4 ECAD tools build on the advances gained by moving from Level 0 Level 1 Level 2 Level 3 to deliver truly advanced design automation functionality. They are the most advanced ECAD tools commercially available today. Level 4 tools often support genuine advances in design process methodology, and hence have the potential to make very significant contributions to solving the electrical and wire harness design problems outlined in the

System integration is an especially difficult and error-prone task, especially as signal count and configuration complexity rises. Imagine a large commercial truck. For marketing reasons these trucks are available with a huge variety of optional systems and many mechanical configurations (different wheelbases, for example). Complicated configuration logic applies: it not possible to have both right and left hand drive simultaneously, but it is possible to have right hand drive and long wheelbase and high end sound system simultaneously. Literally millions of individual electrical configurations are possible. Now imagine keeping track of all possible configurations to ensure that in-line connectors mate correctly, passing signals from one harness to another. This is the task that faces the system integration wiring engineer. In the generative design process system integration and subsequent development of physical wiring is accomplished automatically by

International Wire & Cable Symposium

268

Proceedings of the 56th IWCS

Level 4 ECAD tools from four much simpler inputs. These inputs are (see Figure 3): 1. System pin-to-pin signal connectivity, with each signal being tagged with an expression defining the options it supports. This is a comparatively straightforward design task. Mechanical constraints, typically developed using 3D MCAD systems, such as routing channels, bundle lengths and in-line connector locations. These can also reflect configuration variability such as alternative wheelbases. Configuration logic, which expresses the relationships between the configuration variables, such as the impossibility of building a vehicle that is simultaneously right and left hand drive. Design rules, which implement the mental processes of the system integration engineer.

Select wire part numbers from a library of approved components on the basis of voltage drop calculations, but never smaller than 24 AWG.

2.

Such rule sets can become very sophisticated: they may be hierarchical and dynamic, and may act on data retrieved from outside the CAD data environment (such as the spot price of copper). Rule sets will normally be company or project-specific. Generative design brings many benefits: Designs are correct-by-construction. If the four simpler inputs are correct, the software will automatically generate correct wiring designs for all allowable configurations. Rule sets robustly implement best practice. This allows organizations to capture and develop their competitive intellectual property (rule sets are often written using Level 3 extensibility capabilities). System integration is rapid, accomplishing in a few hours a task that would otherwise take weeks. This not only reduces electrical design times and costs, it also allows alternative designs to be studied (maybe by applying different rule sets) to arrive at designs optimized for cost, bulk, component reuse or other desirable metrics. Design change management is simplified. Because the four inputs are independent, design changes (for example a change of configuration logic) can be rippled through the process easily.

3.

4.

These four independent inputs are all that is required to calculate automatically the wiring for the entire vehicle for the superset of all configurations allowed by the logic. This calculation accomplishes system integration entirely within the ECAD tool.

Figure 3. Generative system integration process The design rules are the key to the process. They capture the skills of the designer, or more broadly the intellectual property of the designing organization, and instruct the ECAD tools algorithms how to proceed. Example rules that could be applied might be: Always place the battery in the front battery tray, except for long wheelbase vehicles, in which case place the battery in the rear tray. Never permit more than six wires in any splice. Always implement safety-critical signals using crosslinked polyethylene wire. Only allow multi-terminations if todays spot price of copper is below a certain value. Never permit wires of the same color to occupy adjacent connector cavities.

Philosophically we can see that the generative process has raised the level of design abstraction: integration engineers now design the rule sets rather than the wiring itself, and allow the ECAD tool to generate wiring using the rules. Raising the level of design abstraction is a well-known technique for coping with design complexity, and there is a perfect analogy here with silicon integrated circuit (IC) design. Early ICs were simple, containing only a few transistors, and each transistor could be designed manually. But as complexity grew this became impossible, and so it became necessary to raise the level of design abstraction. Through a series of abstraction jumps, it is now possible routinely to generate correct IC designs with a billion transistor equivalents from a few hundred lines of C code. This analogy gives confidence that generative design is indeed a key to solving some of todays design complexity problems noted above. But we can also see that generative design is a major design process methodology change, requiring new organizational skill sets and careful deployment management.

3.2 Diagram Generation

A second example of Level 4 design automation is diagram generation: the automatic creation of engineering diagrams directly from underlying design data. Noting the graphical paradigm prevalent in this domain, many different types of electrical and wire harness diagram are typically needed. One example is the set of wiring diagrams to be used by service / maintenance organizations. Service diagrams do not necessarily contain new data: they merely represent existing design data in a new way, for example collecting all power circuits together into a power distribution diagram. Manually drawing new

International Wire & Cable Symposium

269

Proceedings of the 56th IWCS

representations, particularly if design complexity is high, is both time-consuming and error-prone. Two technologies are needed to solve this problem so that multiple related diagrams can be created automatically, both supporting the principle that a diagram is merely a representation of underlying data (this wire connects that pin with that splice). The first is an ability to traverse the design data via queries in order to extract the set of information to be drawn. This is comparatively straightforward, especially as most Level 4 ECAD tools build on the data centric Level 3 architecture: flexible data querying is a natural strength of databases. The second technology needed is much more difficult. The software must be able to render the assembled data into a diagram or series of diagrams that is not only aesthetically pleasing but also must probably adhere to a strict standard. This is a challenging task requiring sophisticated algorithms that can lay out objects in a particular fashion (ladder logic, for example), eliminate overlap, manage white space, respect conditionality (insert this data here if that is true), add intelligent borders, remember edits, and accomplish many other graphical tasks. Subsidiary features such as web-based viewing, diagram markup, language translation and intelligent pagination are also desirable. Some of this graphical technology can be put to a slightly different use. Wire harness manufacturers supplying transportation OEMs are frequently required to develop harness drawings to that OEMs standard. But a major harness manufacturer may serve multiple OEMs, each with their own harness drawing standard. Traditionally such drawings have been developed manually. If software assistance is available the packages are typically hard-coded to produce drawings to a single standard. But by using modern diagram generation technology users can pick from a sophisticated library of graphical standards, automatically producing the correct format. This in turn allows the harness supplier to standardize on a single software package across all customers, which supports the common tools & processes paradigm that can substantially enhance organizational productivity. Despite the difficulty of the task, diagram generation tools are now becoming available. To be sure, technical development of these tools will continue for some time to improve the quality of generated diagrams. But some are already in production use, thus automating another task within the overall wiring design process.

completely eliminate manufacturing errors ultimately traceable to a mis-applied design change. The problem even more difficult when the harness supplier is also responsible for some value-add aspects of harness design. For example the harness supplier may have: authority to move a splice to improve manufacturing ease; authority to make manufacturingrelated wire length adjustments (connector add-on and knock-off) or responsibility for selecting wire colors within a policy set by the OEM customer. The challenge in this instance is to correctly apply design changes originating from the customer without destroying previous value-add work. This challenge is exacerbated if the harness supplier is receiving design data from asynchronous sources: maybe the harness wiring design data from an ECAD tool and harness mechanical design data from an MCAD tool. Level 4 harness engineering tools (see the green box in Figure 1) can solve this problem by interrogating incoming design data and automatically applying configurable change management policy. Each incoming design object (example: wires) and each object attribute (example: wire color) is surveyed for change, and only permitted changes are propagated to the revised harness design. So, for example, the policy may permit propagation of a changed conductor size but block any changes to the color of the revised wire. In essence the design change policy is acting as a data filter: the principle is shown in Figure 4.

Figure 4. Design Change Management Software application of design change policy automates and controls a significant day-to-day challenge for wire harness manufacturers. Further functionality can be added to report the harness design changes, accurately cost the changes, and pass the new data on to factory operations such as wire cutting or automatic test equipment. All these capabilities are provided by modern harness engineering ECAD tools.

3.3 Wire Harness Design Change Control

Most wire harness suppliers face a daily problem: design change. By their nature wire harnesses are physically flexible, and all but the simplest harnesses are still essentially hand built without the aid of much design-specific tooling (maybe only a set of formboards). These two factors mean that it is often comparatively easy to implement a design change (for example increasing a conductor size to overcome an electrical problem or adding a foam spacer to eliminate a rattle) on the wire harness itself, rather than resolve the problem at its root. Furthermore wire harnesses are complex assemblies of components, which are sourced from multiple subsuppliers. Given the number of components in a large wiring harness, specification changes and obsolescence are regular events. This is another cause of design change. Control of wire harness design change is therefore a major problem, causing many hours of validation to be spent which still do not

3.4 Failure Modes & Effects Analysis

A fourth example of Level 4 ECAD functionality is electrical failure modes and effects analysis (FMEA). Irrespective of whether generative or traditional interactive system integration and wiring design is done, there is often a need to test the electrical design against failure. The objective is to improve the design so that failures become less common, and the effects of failure less serious [6]. As an example, imagine a set of electrical systems sharing a common ground point. If this ground point should fail, perhaps all the associated systems will fail. If one of these systems is a safety-critical system, such as automotive adaptive cruise control, the consequences would be catastrophic. FMEA methodology is widely recognized as a key contributor to product safety and reliability, and indeed is a mandatory process in industries such as commercial aerospace. Unfortunately, rigorous

International Wire & Cable Symposium

270

Proceedings of the 56th IWCS

electrical FMEA is difficult to perform robustly, a problem that compounds as design and configuration complexity expands. Level 4 ECAD tools are now available that provide computer assistance to generate rigorous FMEA reports. Electrical FMEA applications build on electrical simulation capabilities (which may be available natively in Level 2 tools or bolted on to Level 1 tools) to probe possible failure events exhaustively. To accomplish this models are built within the software defining the state of each device (a cruise control actuator, for example) in response to all possible electrical inputs. Failure modes (such as the ground point failure) are catalogued and assigned probabilities, and scenarios are built describing the effects of error states (example: actuator fails to respond danger of collision). Severity ratings are assigned to each effect, and detectability metrics also assigned to each failure. With these inputs and explicit knowledge of the electrical design (connectivity) the FMEA software is able to cycle through every possible failure mode, or combinations of failure modes, to generate risk measures against the design. Critically, the FMEA exercise can be scoped to various levels: to an individual system, a group of systems, or (given enough computing power) every configuration of the entire vehicle. These risk measures can be ranked and described in natural language, allowing engineers to take corrective action to improve their design. Although undoubtedly requiring a some up front investment in terms of model building, rigorous FMEA implementation will typically pay back handsomely in terms of avoided warranty and liability costs and sales success via improved brand image due to electrical reliability. In addition to generating rigorous FMEA reports the most modern Level 4 ECAD tools can dynamically identify erroneous connectivity such as sneak circuits (see Figure 5) and even recommend certain types of design improvement to the engineer. Electrical FMEA, design improvement recommendation and connectivity error detection are all excellent examples of ECAD decision support functionality.

with functionality such as versioning, temporal effectivity and cross-organization translation is also critical. The two together are the foundation enabling technology for modern ECAD tools. At this point it is important to differentiate between ECAD tool design data management and large scale corporate PDM systems (sometimes known as Product Lifecycle Management systems). ECAD tools exist to provide design automation in the electrical wiring domain. The data model and associated functionality are specialized to this task. Other design domains are served by equally specialized tools, such as MCAD. In contrast PDM systems are designed to store, control and communicate large amounts of data relating to the total product. As such PDM systems typically do not have sufficiently rich data models, far less the advanced design automation functionality described above, to supplant the specialized tools. It is perhaps helpful to think of the specialized design tools as servants of the PDM system. The specialized design tools receive data from PDM that is correctly held at the platform level (component library data, for example) then release completed designs back to PDM for vaulting and communication. But the creative design task between these events is best accomplished using domain-specific tools. These domain-specific tools may communicate directly with one another, or indeed communicate via the PDM system. This concept is illustrated in Figure 6.

Figure 6. Illustrative PDM / design tool architecture

4. ECAD tool choice and deployment

For many organizations selection and deployment of a new ECAD toolset is a significant undertaking. Level 1 solutions are common, and although such solutions may struggle to cope with todays design challenges, there is sometimes a degree of comfort with the familiar and hence a reluctance to confront growing design issues. It is important to approach the problem from a business perspective, identifying benefits that might accrue so that any investment can be justified and compared against alternative use of resources. Depending on the specific industry, example benefits could include: Figure 5. Dynamic detection of erroneous connectivity Shorter design cycles, leading to reduced time-to-market. Fewer design errors, leading to lower rework costs, fewer quality-related costs or reduced product liability. Improved component management, leading to reduced purchasing costs. Improved wire harness design, leading to lower costs or lighter weight. Enhanced CAD environment, leading to reduced IT costs. Common tool & process opportunity, permitting resource load balancing and enhancing organizational flexibility.

3.5 Design Data Management is Key

Underlying all the Level 4 examples above is a unifying concept: design data management is central to providing the value-creating software functionality. This requires a comprehensive, granular electrical data model, first embodied in Level 1 ECAD tools and extended in scope at Level 2. But the ability to manage this data

International Wire & Cable Symposium

271

Proceedings of the 56th IWCS

Consistent application of best practices / intellectual property, leading to sustained competitive advantage.

To uncover these benefits some form of process and data analysis is needed, not only covering the electrical and wire harness design process itself but also knock-on issues such as warranty costs. Large organizations may choose to employ professional business process consultants who have developed robust methodologies for such studies [7]. But smaller organizations can often make rapid progress without such help, provided a holistic approach is taken. With an analysis in place a new or improved design process can then be developed. It is important that the ECAD software should support the chosen design process, rather than the process be designed on the basis of what the ECAD software can support. The challenges of designing the electrical system for a large passenger aircraft are quite different from that for a lawnmower! It is helpful to bear in mind the five ECAD levels described above. Organizations facing moderate design complexity may well conclude that Level 2 ECAD tools are entirely adequate, while organizations facing higher design complexity will almost certainly benefit from Level 3 or Level 4 tools. But experience has taught some additional lessons: First, ECAD tools should be selected for the long term. It is rare to change ECAD tooling during the course of a transportation platform life: data migration can be quite onerous, especially from Level 0, Level 1 or Level 2 tools. More normally new processes and tools are deployed progressively as new platforms commence. So because transportation platforms are much longer lived than (say) consumer electronics products, ECAD tools may remain in place for ten, twenty or more years. So when making a tool selection it is important to anticipate future technical and business needs as much as possible. Second, data integration with other domains is often a particular challenge. Some organizations select their CAD tooling, and indeed other enterprise systems, on the basis of integration capabilities rather than specific functionality: best to integrate rather than best in class. This approach has definite merits: the IT architecture of Level 3 ECAD tools and design change management capabilities of Level 4 tools are key contributors here. Third, especially where the new design process represents a significant change (employing Level 4 functionality, for example), deployment should proceed in a stepwise, phased fashion. Each phase should be small enough to be realistically achievable, and bring genuine business benefits in its own right. This last point is critical: not only is timely success good for staff morale, it also allows the tool deployment project to be delayed or even cancelled after each phase should the situation change. Business risk is also mitigated by a well-designed phased approach: risk minimization is often a key objective in such process and tool change projects.

It is possible to identify five roughly chronological levels of ECAD software tools, which bring increasing sophistication and automation to the design process. Key to delivering the most advanced functionality is a comprehensive electrical data model coupled with excellent design data management capabilities. It is important to identify which level is most appropriate for a particular organization involved in transportation platform design, recognizing that: New ECAD tools must add concrete business value, and be deployed in a way that minimizes risk. ECAD tools are comparatively long-lived, and hence must be selected with a clear view of likely future needs. Data integration with other domains is frequently a major challenge.

In all but the simplest transportation platform design situations at least Level 2 ECAD tools can bring substantial benefits [8]. In more complex situations, or where design process control or data sharing is critical, Level 3 or Level 4 ECAD tools will be needed. Selection and deployment of new ECAD tools can be a significant undertaking, often requiring executive sponsorship.

6. References
[1] D. Mike, Eye On Electronics,, Motor (May, 2006) [2] As an example, see the case study published at

http://www.emcien.com/EWP%20701%20%20Automotive.pdf.
[3] T. Sedran, T. Wendt & A. Benecchi, Electronics & automotive: achieving a more solid union Automotive Design & Production (May, 2005) [4] Levers for Trouble-free electronics, McKinsey & Company (September, 2004). [5] See http://www.service-architecture.com/index.html for

an introduction to SOA.
[6] See http://www.fmeainfocentre.com/ for an introduction

to FMEA.
[7] See

http://www.mentor.com/products/cabling_harness/succe ss_stories/upload/valeo_case_study.pdf for an example of consultant-led process re-engineering in the wire harness design domain.
[8] N. J. G. Smith, Cutting Costs with COTS, SAE Off Highway Engineering, 38 40 (June, 2007).

5. Conclusions
This paper has described how powerful external forces are causing a discontinuity in transportation platform electrical and electronics design. Specific challenges relating to soaring signal count and configuration complexity are combining with common business drivers to create immense challenges. Among several technologies, Electrical Computer Aided Design (ECAD) software can help meet these challenges.

International Wire & Cable Symposium

272

Proceedings of the 56th IWCS

7. Picture of Author
Dr. Nick Smith, IESD Product Marketing Director, Mentor Graphics. Nick joined Mentor Graphics in 2001, before which he held various positions at Raychem Corporation and Tyco Electronics, including management of both development and product marketing of electrical interconnection components. He has more than 15 years of experience in the electrical / wire harness domain across the automotive, aerospace and rail industries, and has substantial international experience. Educated at Cambridge University, Nick holds a PhD in laser physics and an MBA specialized in high tech marketing. He has authored a number of patents and has published a variety of papers.

International Wire & Cable Symposium

273

Proceedings of the 56th IWCS