Top image is the CAD drawing of a part; bottom is the actually completed part.
Compared with the prevailing practice of converting 2D CAD drawings into static documents, 3D models embedded with data and instructions that computers can parse and process are expected to open the way to significant operational and bottom-line benefits. These include reduced cycle time and cost, less duplication of effort, lower risk of errors, increased part yields and higher-quality products.
Hedburg says the project is based on the idea of model-based manufacturing. “The whole concept is taking some 3D information and moving with it into the digital age,” he says. “We’re looking at 3D CAD models with usable information in them and being able to directly repurpose or reuse the engineering models directly into manufacturing.”
Collaborators in the NIST-led project include interoperability-focused manufacturing services providers International TechneGroup Incorporated (ITI), Milford, Ohio, and Advanced Collaboration Consulting Resources, Summerville, South Carolina.; Rockwell Collins, an Iowa-based manufacturer of avionics and communication equipment for defense and commercial uses; and Geater Machining and Manufacturing, an aerospace supplier located in Independence, Iowa. Other participants are CNC Software, a Tolland, Connecticut, maker of computer-aided manufacturing (CAM) software; Mitutoyo America, a maker of measurement equipment and software; and software vendor CoreTechnologie, Rossford, Ohio.
Leveraging a New Standard
The Design to Manufacturing and Inspection Project leverages a new international standard for incorporating computer-readable product and manufacturing information (PMI) into 3D models that do not require human interpretation of graphical depictions and then, data reentry. Just published by the International Organization for Standardization, ISO 10303-242 (best known as STEP AP 242) enables designers and process and systems engineers to embed 3D representations of parts with actionable specifications for materials, geometrical and dimensional tolerances, and surface texture, as well as process notes, finish requirements and other information.
Like other interoperability standards in the STEP family, AP 242 serves as a mediator between programs from different vendors. For example, a part design prepared with proprietary software is converted into the common language, format and protocols of STEP AP 242 so it can be read by another vendor's software, which typically retranslates the design instructions into its native language.
In the new project, Rockwell Collins will use its CAD system to generate a 3D design of a part, complete with all feature tolerances and other specifications. The design will be translated into STEP AP 242 so that Geater can repurpose the model into the language understood by the software it uses to generate machining instructions. Separately, Geater will reuse the STEPAP 242 model in software—with no manual data entry—to generate code that will direct a coordinate measuring machine to perform the inspection measurements necessary to determine whether the part is manufactured as designed.
At each stage in this thread, researchers will verify and validate translations involved in the data exchanges.
The real-world problem the project addresses is that while seamless data exchange is simple enough if everyone is using the same CAD system. However, this approach locks a company into a single proprietary solution, and it is not practical for companies to exchange product and manufacturing information with many suppliers and customers.
“A portion of this [data exchange] can be done today if you have the same CAD system on both ends,” says Hedburg. “We’re trying to close the gap by bringing in neutral formats and extensions of embedding requirements so disparate systems can work together.”
He continues, “We see this project as leverage some of these technologies to integrate all of this into a collection of capabilities that can be realized. AP242 has ability to bridge disparate CAD systems. Also what is missing in these models is the exact information requirements so that everyone else in the lifecycle can complete their task. We’re also developing standard information models to support entire lifecycle.”
A full-scale demonstration of end-to-end interoperability is expected by summer 2015. In follow-on work, NIST researchers will focus on strengthening and extending the digital thread to link a broader range of manufacturing-related activities such as assembly, bidding, engineering changes, in-process inspection, and visualization and collaboration.
