中文 English

A New Method For Electrical Systems Design

Adopting a model-based approach to seamlessly leverage data from development stages to drive downstream processes.

popularity

Electrical system complexity is reaching a tipping point across industries, from modern passenger vehicles to sophisticated industrial machines that can now contain nearly 5,000 wiring harnesses. The electrical systems of these machines contain multiple networks, thousands of sensors and actuators, miles of wiring and tens of thousands of discrete components (figure 1). Designing these complex systems is a significant challenge itself, but there also needs to be understanding of how system requirements move into manufacturing and maintenance processes.


Fig. 1: Vehicle electrical systems are becoming immensely complex, containing thousands of components and miles of wiring.

Capitalizing on these industry shifts requires change. Producing the next generation of products will require meticulous traceability and improved organization. For example, automotive manufacturers no longer design a car and produce millions of copies. They design one of a billion, with unique vehicles rolling off the production line with unique configurations. EDS designs need to support such numerous vehicle configurations while minimizing costs. Traditional EDS engineering methodologies in use across the automotive, aerospace and off-highway industries are ill-equipped to manage this level of complexity. In particular, traditional engineering methodologies fall short in three ways: they create silos between engineering domains and abstractions, rely on labor-intensive manual processes, and lack robust data continuity throughout the development flow.

The electrical design flow

When tasked with designing an EDS and wiring harness for a new vehicle, there are four main stages to development for an electrical engineer. These stages are the definition of the electrical and electronic (E/E) architecture, followed by detailed design in each domain, creation of manufacturing documentation and, finally, the creation of service and maintenance documentation.

In a traditional engineering flow, these development stages are connected only through ad hoc interactions and manual transfers of engineering data or drawings. The various electrical domains and other engineering disciplines, such as mechanical engineering and software development, operate in silos with little to no visibility of the actions of other engineering teams. As a result, the traceability of data from vehicle requirements to the functional abstraction and down to the eventual implementation in the wire harness is weak or entirely absent.

In contrast, a primary benefit of adopting a model-based approach is the ability to seamlessly leverage data from each of these stages to drive downstream processes. The architectural definition can be used by network, software, electrical and harness engineers to derive the necessary inputs to begin detailed work in each domain. Likewise, the detailed design from each electrical domain can be used, enriched and built upon for each downstream area and can be used to generate manufacturing and service documentation. A robust digital thread means that engineers no longer have to exchange this data manually, improving accuracy and ensuring the traceability of data throughout the development flow. The implementation of a model-based flow with an advanced electrical systems engineering solution brings additional automation, analysis and multi-domain collaboration capabilities.

Generative design: The heart of a model-based approach

Generative design is critical to extracting the most out of a model-based development methodology. Generative design takes inputs from a variety of sources and automatically fuses them together to produce the required output for that stage, which can also be the input required for the next stage of development. To generate the EDS, the inputs will include the physical harness topology from the mechanical CAD (MCAD) environment, the product plan, logical system designs and a list of existing parts to reference.

Each of these inputs provides certain constraints on the potential output of the generative design. The harness topology from the MCAD environment provides physical constraints for the wiring and other harness components, while the product plan inputs engineering and marketing constraints on vehicle features and options and how they can be combined. Generative design then synthesizes wiring, connectors, splices and more to satisfy the required connectivity, physical dimensions and other constraints.

The true power of generative design is the speed with which it can synthesize optimized designs from the massively complex constraints of modern vehicles. For a complex system, synthesis may consider hundreds of millions of potential implementations, eliminating those that fail to meet requirements, and do so in a fraction of the time it would take a human engineer.

However, generative design can still produce millions of possible EDS implementations even with the design limitations and restrictions from the mechanical design, logical systems and other inputs. Additional refinement is required. The most effective means of further narrowing the design space is through the intellectual property (IP) engineers gather over years of experience. Vehicle manufacturers that employ generative design can capture this company IP and engineering experience in design rules that guide synthesis, integrating the IP into the EDS or other generated designs. Using design rules, companies not only preserve the knowledge and experience their engineers have built up over decades of work, but can also improve the results of generative design by further limiting the acceptable outcomes.

Optimizing variants, complexity and cost

Today, OEMs are not synthesizing just one EDS or vehicle design. It is now standard for OEMs to offer catalogs of optional features and to produce different variants for different regional markets. The result is billions of possible unique vehicle configurations. OEMs employ a number of strategies, such as bundling features together, to reduce this number of possible vehicle configurations. One of these is to make certain wiring or other harness components standard on a range of derivatives, even if the vehicle they will support does not use these wires or components. By giving away content in this manner, OEMs increase the part cost of each harness produced, but can reduce the logistical cost by eliminating unique harness part numbers that need to be managed (figure 2).


Fig. 2: Reducing the number of vehicle derivatives that needs to be managed may increase individual part cost, but also reduces logistical cost and complexity.

Even so, OEMs still manage thousands of vehicle variants across several regions. Manual complexity and cost calculations that are often based on assumed, ‘good enough’ values will not suffice in the future. A model-based flow provides data, rather than assumptions, on which to base decisions around harness complexity and cost. Vehicle models can be analyzed to determine what content should be given away to achieve a balance between cost and complexity.

The rapid design iteration and robust analysis of model-based flows can also support changes in the business or manufacturing model used by an OEM. With generative design and a model-based flow, engineers can simply tell the E/E systems development software which business or manufacturing model to use (such as a composite/derivative or modular harness assembly method) and the software will generate models and other outputs to match. The engineers can perform further analyses, and even compare between business models to select the best option. The robust metrics provided by E/E systems development software then enable deeper understanding of the designs, creating a feedback loop of design generation and evaluation. These may include annualized cost projections, weight, and more to drive correct decisions on both the harness implementation and business model used in production.

Conclusion

Achieving the engineering efficiency, excellence, and speed needed to bring innovative products to market requires a methodology that promotes collaboration, automates processes, and features robust traceability. Model-based approaches, implemented with an advanced portfolio of engineering software, provides these capabilities. They create an unbroken digital thread from vehicle requirements to functions and all the way to physical components on the vehicle. Such robust traceability ensures that vehicles leave production fully verified and validated, meeting all requirements from the manufacturer, customers, and governing authorities.

Furthermore, by implementing the model-based approach within an advanced electrical systems engineering solution, manufacturers of complex machinery can leverage generative design and robust metrics to rapidly create and assess designs. Within this flow, engineers can use generative design to integrate models from various domains and produce an optimized output for downstream processes. These capabilities also support the optimization of harness cost and complexity associated with vehicle variants. As products continue to become more complex, the traceability and accelerated development cycles provided by this model-based approach and generative design will prove crucial to success.



Leave a Reply


(Note: This name will be displayed publicly)