The automotive industry is undergoing its most important transition since the introduction of assembly robotics and automated work cells. The interconnected world of personal devices and Internet of Things (IoT) is already present in vehicle design. Greater factory automation and learning systems are just a corner or two away. For the automotive design, manufacturing, and systems engineer, the quality of the digital master model that will transform and command these interdisciplinary functions is critical.
Data must be as unambiguous as a paper 2D drawing once was for machining and as chock-full of product information as the file cabinet in a director’s office – only now it must be instantly accessible globally to machines awaiting instruction or a computational fluids analyst studying an injection, exhaust, or brake system.
Digital design in automotive
The automotive industry still rides on uneven pavement in the digital world. Its centerpiece for containing product information, the computer-aided design (CAD) system, derives largely from several key vendors with proprietary math kernels defining shapes and surfaces. Third-party contributors of specialized software add to the different recipes, which do not communicate uniformly and in agreement, despite advances in STandard for the Exchange of Product model data (STEP) and other neutral file formats. Small human errors or digital discrepancies in edges, faces, tolerances, or draft angles often remain, interfering with everything from a finite element analysis (FEA) to a press operation.
Why does digital quality matter in an industry that has succeeded by navigating its own legacy tools and practices? Global competition increases annually, making no geography or brand safe from breakthroughs in performance, cost, and appeal. Trends sweep in quickly, such as the move to SUVs from cars. Whole platforms, assemblies, and exteriors may need to change instantly to accommodate new markets or external requirements. More accurate, reliable digital systems can quickly take manufacturers down new roads.
Design intent, virtual testing, and validation must travel via the CAD model to downstream production and inspection stations and back again to design for continuous improvement. The one central element that all these systems must share is accuracy and readability.
Twists in the road
Unlike aerospace, which at this moment leads automotive in establishing digital uniformity, vehicle makers have significantly tighter financial margins, shorter product life cycles, shorter development cycles, and fewer internal and external resources (such as NASA or DARPA) to explore model-based definition (MBD) and the model-based enterprise (MBE). Aerospace leaders have mandated MBD for at least their Tier 1 suppliers. In automotive, Tier 1s seem to be responsible for bringing into agreement data from their sub-tier suppliers and the original equipment manufacturers (OEMs). Being in the middle of the data hub often forces the Tier 1 group to lead data integration by example, enforcing best practices by default.
Since all OEMs have their own set of modeling requirements as deliverables from suppliers, Tier 1s must replicate each OEM CAD environment to achieve acceptable delivery. These requirements often include different CAD formats, software versions, and flavors that are specific to each OEM as well as individualized rules about geometry quality, metadata, usage of layers for product manufacturing information (PMI), and delivery mechanisms. Tier 1s manage more than 30 CAD environments – and those have different versions and practices within vendors.
When Tier 1s reuse core knowledge that goes into their products to create custom variations, they must manage different OEM environments and update their ongoing changes into each system throughout the engineering journey, without losing fidelity. At the same time, they must adapt and pass different sets of quality criteria.
Manufacturers don’t want engineers spending time to manually solve these oversight functions. Engineers typically want to focus on the design, test, and production, not coding nuances within CAD/computer-aded manufacturing (CAM)/computer-aded engineering (CAE) that manifest problems in hard-to-determine areas. Yet these code-related issues can block precision machining and cause rework and delays in every discipline.
The solution is in independent quality and translation software systems – driven by native application programming interfaces (APIs) unique to each CAD program – that bring supplier geometry and the digital package of annotations and manufacturing instructions into a target master model of the OEM. Such automated back-end processing can overcome costly manual repairs, interruptions, and uncertainty.
Suppliers want to solve model discrepancies through a seamless conversion between these 30-plus environments, including unifying underlying data points sent to neutral formats such as STEP and IGES. Many manufacturing and engineering programs are designed to consume these standards of exchange, such as CNC toolpath generation and Parasolid translation for analysis software. The key is a conversion platform that can be easily executed and produce a reliable output so that the end user can immediately perform intended work.
Another goal is doing a preliminary product data quality (PDQ) check before information has been submitted to a supplier, having the toolset to customize PDQ checks to accommodate multiple profiles, and the capability to apply checks to CAD output or a neutral format.
Ideally, a Tier 1 can standardize on a single platform, optimize all products in a primary CAD environment, and simply convert to any necessary target system, conducting a PDQ check before the data is submitted to a customer. Quality problems can be pre-detected and addressed internally in a more automated process, and the model becomes ready for delivery with confidence that it will pass. This approach keeps the OEM from having to reject models, unravel the issues of an unfamiliar system, and re-iterate the process throughout days and sometimes weeks.
Automotive components and systems can generate more than 100 revisions, and required migration steps to new CAD software systems create additional pain points. MBD can support producing vehicles more quickly and less expensively than before. Factory automation must be supported by accurate, all-digital platforms that create clean, quick pathways from CAD to machine systems to inspection programs, and back again to the original CAD model for confirmation of end quality. Advanced healing, validation, translation, and PDQ technology can smooth this vision of automated digital design and production.
Elysium Inc. https://elysiuminc.com