The Quality Information Framework (QIF) Version 2.1 suite of quality information models is an ANSI-approved standard developed by the Dimensional Metrology Standards Consortium. All QIF information models are written in the XML Schema Definition Language (XSDL). Data files (called instance files) conforming to the QIF models are also written in XML. QIF, along with the Dimensional Measuring Interface Specification (DMIS), provide rich and comprehensive digital modeling specifications for every key element in the manufacturing quality measurement process, from digital part models, to measurement resources, to measurement rules, to measurement plans, to measurement programs, to measurement results, to summary measurement statistics. Measurement programs are generated by measurement plans, which in turn are generated from part models and a set of measurement rules. Measurement rules are intended to help users select appropriate measuring equipment, fitting algorithms, and methods for reporting measurement results. Since the use of QIF in actual software systems is in its early stages, measurement system software vendors currently create and consume relatively simplistic rule sets and plans. Furthermore, the ANSI QIF Rules 2.1 standard is still in the early phases of development. What is needed now are sophisticated, human-interactive software tools, which can receive QIF-formatted models and QIF-formatted rules as input, to create complex measurement plans which are validated and verified to be correct, complete, and compliant to ANSI QIF. Such a tool would also be used in testing the validity of the QIF specification itself, which is of special importance to NIST. A well designed tool could reduce measurement plan size substantially, depending on the part to be measured and its features.
This subtopic seeks to enable the development and testing of the QIF Rules specification. The awardee will be able to provide a testing ground for the current version of the QIF Rules specification, and provide valuable input for the development of the next version of QIF Rules. The same can be said for QIF Plans. QIF Plans have not been thoroughly exercised but current software implementations of QIF. If QIF Plans is going to be ready for prime time in the dimensional metrology software world, it needs to be tested against more complex requirements and manufacturing use cases.
Phase I expected results:
The awardee will design a structure for the interactive software tool, develop a prototype, and test the prototype with a real part. The awardee will design, create and test the software tool prototype with relatively simple inputs, but all in QIF format. The software tool will demonstrate its capability with a relatively simple part with geometry, features, and tolerances modeled in QIF Model Based Definition (MBD). The tool will process the part, using simplified QIF Rules, and generate QIF-formatted measurement plans. The goal in this phase is to design a tool, and build a prototype, that can be expanded to produce non-verbose QIF-compliant plans for a rich set of part types with a complex and realistic set of features and tolerances, using the specified rules in QIF Plans format.
Phase II expected results:
The awardee will build and demonstrate a more sophisticated measurement plan generation tool. First would be a test of the optimization of the tool against 1) a wide variety of parts and features and tolerances and 2) a wide variety of rules for measurement equipment types, measurement tool types, and feature-type measurement requirements. Second the awardee will test the correctness of the tool against 1) expert (human) analysis of the resulting plans from the tool and 2) the performance of the plan in simulation tools and a real environment. For example, the rules portion of the interactive software tool might have three interrelated functional modules: a module to allow intelligent selection of coordinate machines and probes, a module to allow intelligent selection of fitting algorithms for processing measured points to fit substitute features and a module to allow intelligent selection of the methods to report measurement results. During the software development, information models for software engineering have to be QIF-compliant, e.g., for product models, resource models, and rules models.
Finally, the interactive software tool will generate programs in DMIS for coordinate measuring machines (CMMs) from QIF-formatted plans. The system will also be able to accept a QIF Resources file as input. The proposal may be to extend an existing CMM programming system or to build one from the ground up.
NIST researchers may be available to collaborate with and provide consultation to the awardee.