Who has a stake in measurement results? Whether you are a participant in manufacturing part of a toaster, a golf club, a jet engine component, or all of them, at the end of the day when a manufacturing process is complete, you want to know one thing; does the part meet all the dimensional requirements?
Whether you are a machinist, a process engineer, a manufacturing engineer, a design engineer, a quality engineer, a data integrator, a company owner, a manager, or a customer, you have a stake in the measurement results.
Since this is CMM Quarterly we can assume that you all are familiar with coordinate measuring machines. So, let’s consider the genre of CMM’s as measurement devices for a moment. To develop the point, let’s suppose you have several different CMM’s with different inspection software programs to run inspections. Do they output the identical measurement results format? Emphatically, you replied “no!”. Of course not, as each CMM brand has its own software to run inspections they also have their own software to generate results. There currently is no standard that defines how measurement data results should be output. Without a standard, it makes it difficult to have two systems report the results in exactly the same way and inevitably this job is left for the end user to do. Wouldn’t it be nice to have a standard?
Let’s assume you worked hard on defining a common measurement results format for your company, it looks good. Now, it’s time to step back and review your work. Can every person in every role get the data they need to fully understand the measurement results for their specific needs? Here are some common questions that might be asked. What tolerances pass and fail? Does the part conform to the applied Geometric Dimensioning and Tolerancing (GD&T) requirements? As long as all tolerances are passing, you are manufacturing good parts and everyone is happy. But then it happens. A tolerance fails, and continues to fail part after part, production is stopped. Machinists and you the manufacturing engineer are reviewing the measurement results report looking to answer questions. What tolerance or tolerances failed? Why did the tolerances fail? How do I evaluate the measurement results? Is it the CMM, the sensor, the fixturing, the CNC machine? Can I fix the part? How do I fix the part? It is readily apparent now that more information is needed than what is available in the common report.
Pressure is on and the need to identify, diagnose and remedy potentially complex dimensional problems is at hand. Production is down. You now realize you have a problem with several complex surfaces on your casting. Surface profile tolerances on several of the complex surfaces are failing. It’s not a simple difference between the surfaces and the referenced datums. It’s time to get the CMM programmer to add statements to the program to get the actual 3D measured data output to a file. Then you have to get a CAD guru to get design points from the inspection program then add them to the CAD model. Several hours later, the CMM programmer has made the changes to the inspection program to output the actual X, Y, Z points. The data file with the measured points is delivered to the CAD guru after the newly modified inspection program has been run. Production is still down. The CAD expert must best fit the point data on the datums to define the measured datum reference frame(DRF), then apply this same best fit DRF transformation to the nominal data points (something that is often overlooked by many engineers that don’t have a good understanding of the mechanics of applied GD&T). Next, the measured point data must be transformed into the proper CAD space. Then finally, the measured points can be evaluated by checking distances between DRF transformed nominal along the surface normal direction (note this deviation is NOT calculated via a simple 3D distance between points another common mistake). The CEO peeks in to see how things are going. The CAD guru and manufacturing engineer work together to diagnose the dimensional problem. They find that one of the locating pins on the machining fixture came out of adjustment causing the part to twist from its nominal machining position. Ok, the problem has been found and fixed; the CEO is relieved and you are glad you still have a job. Wouldn’t it have been nice to import the measurement results data right into the cad model so it could be evaluated without all the steps above?
You are ready to submit the measurement results to your customer. Purchasing finds out at the last minute that the customer needs the measurement data put into their special measurement results format, and want to see not only the few first parts you have produced, but want to have some statistical process control (SPC) data to support your manufacturing capability as well. Now what… Well its going to take a software programmer and perhaps an intern to translate the existing measurement results into the format the customer requires, an SPC program needs to be purchased and the data needs to be extracted from the report format and put into a data format by which the SPC software can be populated. Wouldn’t it be nice if there was a standard measurement results file that could be sent to the customer, or used directly by any SPC program to generate any report needed, and evaluate the manufacturing process?
We have illustrated the big frustrating, cost intensive non-interoperable measurement data results picture by identifying a wide array of stakeholders and their respective levels of data needed for their particular roles. We noted different CMM vendors have different software which generate different and sometimes proprietary measurement results formats. We understand that costs associated with training to understand these formats can continue to accrue. Further, we outlined the complexity of a costly diagnostic process with point transformations by engineering and CAD experts by which dimensional measurement results are used to diagnose and remedy a real-life manufacturing problems. Measurement data stakeholders can go beyond our own brick and mortar walls and implicate a multitude of customers, or even our own supply chains. When considering this problem diagnostics scenario with its complexity and resource intensive requirements, the need for data translation from one format to another to meet everyone’s formatting requirements, translating data for measurement data consumers such as SPC programs, it is easy to see that enormous time, and resource costs can mount. When taking all this into account the problem of measurement results interoperability becomes even more important problem to solve.
Many of us have lived and unfortunately still live this measurement data interoperability nightmare. If you note all this complexity, work, cost and training are all centered on one problem, the lack of interoperability in hauling measurement results data.
Wouldn’t it be nice if a single measurement data results model existed and contained all this measurement information needed, from nominal feature definitions to measured feature information, nominal tolerance definitions to measured tolerance information, datums, datum reference frames, traceability information, and enough information to reanalyze the tolerances? Wouldn’t it be nice if the data model had the wherewithal to store discrete nominal and measured point data in a single reference frame eliminating the need to transform the points to a common coordinate system? Wouldn’t it be great if the populated data model could be used to generate a myriad of different reports, and graphical results, and be portable between otherwise disparate systems? Wouldn’t it be nice if the measurement results could be directly consumed by many commercially available systems such as SPC programs? Wouldn’t it be nice if we didn’t have to export and import, map data elements from one system to another, or post process results from one format to another. Wouldn’t it be nice if there was a Measurement Results Standard?
Fortunately, for all of us a data model like this already exists! And as a bonus, it addresses all the aforementioned “Wouldn’t it be nice if…” items. It is known as Dimensional Markup Language or DML.
We are Validation Technologies Inc. (Val Tech) and Metrology Integrators. This year we have joined the Dimensional Metrology Standards Consortium (DMSC) a consortium of businesses and experts representing the whole of industry organized to further improve the DML data model to make it more extensible, improve integration, and make it more data friendly for both DML generators such as CMM software and DML consumers such as DMLToolkit or SPC software. We are active members and participants in the DMLI working committee of the DMSC. As with any undertaking with the DMSC the end result will be the progression of the DML specification to a full fledged “American” ANSI Standard and then to an ISO standard. With the DMSC we can make a serious stance to make DML a standard to benefit all stakeholders of measurement results data, ease implementations and integrations, and eliminate extreme costs of training and translating measurement results data between people, companies and systems.
As CMM programming consultants, software developers and metrology integrators, we have a lot of experience with the use of DML over the years as a hauler of measurement data. In fact, we are “all in” with the ideals of all the metrology interoperability standards such as DMIS, DML, I++. In fact we have created our own DML consumer known as DMLToolkit. DMLToolkit provides a simple user interface to graphically display inspection results. It allows users to review tolerance results in just seconds and drill down to more discrete tolerance information if needed by placing the mouse over the tolerance bar. With the ability to group related tolerances it can help them quickly find the root causes and problematic manufacturing operations.
In fact, our most resent accomplishments include the integration and implementation of our Inspection Toolkit Suite (ITS) of software which includes DMLToolkit in the worlds largest InspectIt! / Siemens Inspector and DMLToolkit in a shopfloor CMM inspection environment at GE Aviation, Madisonville. Our software was designed and proven to provide a stable shop hardened inspection execution platform to reduce if not eliminate shopfloor support through ease of use and automation. Isn’t it time your company stepped in to join forces to address this interoperability problem? Contact the DMSC at www.dmisstandards.org or us at www.validation.com, or www.metrologyintegrators.com for more information about DML.
Validation Technologies Inc (Val Tech) and Metrology Integrators are software solutions partners with Siemens PLM Automation and Drives, and have been in the business of CMM programming and quality systems integration for over 14 years. Together Val Tech and Metrology Integrators provide the whole package, from specialized Siemens eMPower Quality training specific to your companies applications, to the design, development and implementation of production hardened CMM inspection software applications tailored for your company and supply chain. We realize a rock solid system is required to run in production dependant environments, and it is also important that the system is scalable, interoperable, and able to integrate with upstream data producers and downstream data consumers to provide a seamless convergent quality system. Val Tech has a solid reputation in the industry and has successfully designed, developed, implemented and supported several production related applications.
By Scott Hoffman, Metrology Integrators