By David H. Genest

It's frustrating, like trying to fit the proverbial square peg into a round hole, but it happens to everyone who uses metrology in manufacturing operations. The incompatibility between metrology systems of different manufacturers is a serious challenge for the science of metrology as it enters a new millennium. For decades, metrology equipment manufacturers sold turnkey packages, including hardware, software, controllers, sensors, parts and service. All of these turnkey packages used proprietary standards that forced users to go back to a specific metrology equipment vendor for additional systems and upgrades to existing equipment and software. Perhaps that wasn't all bad if the user had a small shop with a single production operation, because the economic benefits of dealing with a single vendor can be significant, but most users have several production operations spread over a large area or in separate locations. It's in these situations that multiple systems from different equipment manufacturers create difficulties that affect not only quality control but also product development and production.

 In recent years, metrology equipment manufacturers have begun opening their systems with drivers that allow third-party software developers entrée to the machine control function. While this has been a step in the right direction, each hardware vendor developed a proprietary open communications protocol that made it necessary for third-party developers to support not only multiple drivers for each brand, but in most cases, separate protocols for each configuration within the brand. Today, there's a new awareness of system compatibility that is uniting the resources of users, government, trade associations, industry and universities with hopes of developing and refining a standard interface protocol that frees metrology from the restrictions of proprietary operating systems.

Understanding the problem

 Incompatibility issues are more than just a nuisance. When System A doesn't work with System B, users have to stock more than one set of spare parts, train operators on multiple software packages, and produce different types and formats of inspection reports.

 While these problems are significant, the real difficulty that system incompatibility creates is the inability to exchange a common inspection model between measuring devices. The time it takes to reprogram each device to inspect the same part, refixturing time and cost, and the resulting loss of accuracy in the measurement and inspection process creates serious inefficiencies in a system that is, theoretically, designed to provide a high level of process control and improved inspection throughput.

 These data exchange roadblocks also create time-to-market problems. During the product development and introduction cycle, unnecessary time spent reprogramming inspection routines and re-evaluating data contributes to lengthy product development cycles with a resulting loss of competitiveness.

 The effects of system incompatibility are growing rapidly. It has been estimated that the problem has cost the automotive industry alone about $1 billion. Cost penalties also exist for metrology system suppliers and individual users who absorb the costs of repair and training necessary when incompatible systems must work together.

 What is needed is a standard for communication protocols that allows metrology systems from different manufacturers to work together seamlessly. And it's on the way.

An old idea with a new twist

 Common exchange standards aren't new to the metrology industry. The first metrology standard was developed for the exchange of programming information. The Dimensional Measurement Interface Standard (DMIS) allows programs to be shared between different pieces of measurement equipment. However, the problem of incompatible measurement and inspection software still exists.

 A typical coordinate metrology system includes a measurement and inspection software package containing many inputs and outputs to aid the user in programming and controlling the machine. Inputs include a computer-aided design (CAD) interface, offline programming, a graphical results interface, a sensor interface and a machine control interface. In the past, each of these input interfaces represented an opportunity for a metrology system manufacturer to create a proprietary package, although some common standards have been developed.

 For example, there are standards for CAD and offline programming interfaces. The CAD interface is used to transfer CAD data to the CMM for efficient online programming, as well as to produce graphical reports of the CMM output. Several standards exist for the transfer of CAD data. In the United States, the most widely used standard is IGES, while in Europe, VDA is the most popular. STEP is a developing standard for CAD data exchange.

 The offline programming interface transfers a program from an offline programming package to an online package or from machine to machine. DMIS is the most popular standard for this type of interoperability.

 As manufacturing tolerances become tighter, inspection results become more critical. Consequently, an important issue related to CAD interfaces is the need for a seamless, nontranslated flow of information from CAD tolerancing data through offline programming, measurement execution and analysis of measurement results.

 To help meet this requirement for higher accuracy, Brown & Sharpe, a Rhode Island-based metrology company, has released PC-DMIS v. 3.2 with a DIRECT-CAD interface to such programs as CATIA; Unigraphics; ProEngineer; SDRC; and the common CAD format, ACIS. Direct CAD interfaces eliminate the need for the software to translate the original CAD model in any way, ensuring that accurate design data is used for part programming. Users can create part programs directly on the CAD model, reducing programming time and improving accuracy.

 The graphical results-analysis interface communicates measured results to a graphical reporting package to show the user what was measured on the equipment. There isn't currently a standard for this interface, but STEP AP219 is being modified to standardize it. AP219 speci-fies information requirements for representing, exchanging and archiving information required to perform dimensional inspection.

 The inspection machine controller interface would standardize the communication between the inspection software and the machine controller. One approach has been a user consortium under the auspices of the Metrology Automation Association (MAA) that has focused on the development of a common driver for interface between machine instructions and task programs, a coordinated testing method and a common reporting interchange. The result is the Metrology Common Driver (MCD), a set of common standards that will reside between the user software and the hardware controller, allowing the hardware to be controlled by any software that supports the MCD. Currently, the MCD is concentrated on computer numeric controlled coordinate measuring machines, but it will be expanded to cover all types of metrology hardware.

 Standards development of this sort generally takes an extremely long time. To help develop a useable protocol quickly, Brown & Sharpe teamed with Carl Zeiss IMT Corp. and LK Metrology Systems Inc. to create a common driver, establish commands that will be common to each company's machines and agree on a syntax that will work for all machines.

 The standard is under review by the National Institute of Standards and Technology. NIST's initial report indicates that the concept has merit, and work is underway to launch a full-scale development project at Quality Expo International in Chicago in April 2001.

Our common future

 Working closely with NIST, the MAA is assisting in developing an industry-government partnership to create a National Metrology Test Bed. The purpose of the NMTB is to specify, implement and test a set of interfaces that will allow a smooth flow of information between each of the major steps in developing and executing automated measurements. The test bed may consist of equipment and software at a number of industry sites, as well as the NIST laboratories. With the test bed, researchers and users can test new developments and compare them with known results of other research in the same field.

 Also involved in this industrywide effort to improve operations and data exchange is the University of North Carolina, Charlotte. UNCC, which has one of the most advanced metrology education programs in the United States, will conduct a two-year research project to identify issues related to integrated metrology, develop prototype systems and contribute to NMTB demonstrations and technical assessments.

 This approach to integrated metrology operations will offer users significant gains in productivity, throughput and efficiency. It's truly the dawn of a new era in measurement and inspection technology. And integration, with its potential to harness the power of dimensional measurement for improved process control and product quality and reliability, is an idea whose time has come.

References

1. Clifford, Michael. "Metrology Common Driver," Ford Motor Co.

2. "Update on Innovation Initiatives in Metrology Automation," Metrology Automation Association Web site, www.metrology automation.org.

3. Waite, Robert D. "Need for a National Metrology Test Bed," DaimlerChrysler Corp., September 2000.

About the author

 David H. Genest is director of marketing and corporate communications at Brown & Sharpe Manufacturing Co. in North Kingstown, Rhode Island. Genest is a member of the Metrology Automation Association board of directors. E-mail him at dgenest@qualitydigest.com .