Re-Visioning Medical Part Metrology

Medical part manufacturers worldwide are scrambling to take advantage of recent advances in vision/multi-sensor measurement machines, new sensor technologies and cross-platform, CAD-based software. In both product development and production manufacturing settings, these tools are providing their early adopters with a greatly expanded set of measurement capabilities. These include:

• Using 3D vision/multi-sensor measurement systems to perform geometric and dimensional analyses (as opposed to simple, 2D functional measurements) of sur faces or geometries that are too small, too flexible or inaccessible for conventional tactile measurement.
• Using powerful, CAD-based, metrology programming software to dramatically reduce part program backlogs and keep pace with aggressive product development cycles.
• Using multi-sensor equipment to improve measurement accuracy while at the same time relying on single setup capabilities to increase measurement throughput.
• Using cross-platform metrology software to reduce training costs, while giving the users access to a substant ial ly expanded set of measurement capabilities from within a single programming environment.
• Using a wide range of data collection and data management functions to access powerful programming and analytical tools previously unavailable to vision and multi-sensor systems.

Measurement Hardware

Major surgical products and medical device manufacturers have been investing heavily in new multi-sensor measurement systems due to their advantages in shortening product development cycles, improving part validation throughput, and reducing rejection rates for parts with complex geometries or extensive sculptured surfaces. These manufacturers are also encouraging, if not insisting, that their suppliers adopt similar approaches.

The CMM vision probe, mounted on an articulated wrist, allows the software to orient the probe to the part's surface for precise measurement.

The strategies they are using to acquire advanced 3D vision/multisensor measurement capabilities are diverse:

• Purchasing new, stage-type vision systems with standard 3D CAD-based vision software.
• Retrofitting older, popular models of stage-type vision systems with 3D vision software.
• Acquiring new multi-sensor measurement systems (including those with dual Z-axes for more efficient utilization of both probes and cameras on the same machine.)
• Using new camera probes that are compatible with the articulating wrists used on their coordinate measuring machines (CMMs).

While there are many choices, the measurement equipment itself is no longer the primary concern of today's users. Instead, they are more focused on determining sensing technology or technologies best suited to the specific application. The hardware follows from that determination.

Probing Options

Whatever the part measurement objective, there now exists a sensor or combination of sensors that can accomplish the goal effectively and productively.

Vision and multi-sensor measurement systems allow users to create measurement programs.

Tactile probes: Touch probes are the fastest and most accurate means of quickly collecting high-resolution point data over wide stretches of medical part territory. Many times, though, these types of parts include some features that a touch probe cannot measure effectively. This might be because the features are too small for the probe, or inaccessible to it, or would be deformed by it.

Measuring these parts correctly demands a solution that allows probes of different types to work together on the same machine – a multi-sensor solution. In these sorts of systems, the idea is to employ each sensor to its best advantage. For a tactile probe, a typical scenario would be to first crosscalibrate it with the other sensors on the machine and then use it to locate the part accurately on the machine using iterative alignment methodology. With those tasks done, the user can then confidently measure each feature with the appropriate probe. There are a growing number of sensor types that are now usable in combination with tactile probes, as well as in stand-alone applications.

Cameras: Once, cameras were only appropriate for 2D functional (go/no-go) measurements. Now, dramatic improvements in focusing ability and edge detection algorithms mean that cameras can accurately collect 3D data and compare it directly to the CAD model using cross-platform metrology software. Additionally, these software packages provide flexible control over the variables critical to effective vision measurement, including parameters like focus, lighting mode, lighting intensity and edge detection variables. Within a narrow field of view, cameras are capable of instantly collecting large numbers of data points for variable data analysis.

Micro-optical probes: While there are many applications in which medical part manufacturers would prefer to use tactile probes, the sheer smallness of many holes, slots and other features prevent the probe's entry into these tight spaces. Micro-optical probing dramatically reduces the size of the tactile probe by eliminating all of the electronic and mechanical paraphernalia associated with the touch-trigger mechanism. Instead, a camera trained on the probe tip senses deflection, thereby indicating contact of the probe with the part's surface. This technology makes tactile probing usable on features many times smaller than previously possible with the tiniest of touch trigger probes.

White light sensors: This technology analyzes the wavelength of returned light to provide resolutions measured in nanometers (hundredths of a micron). One major advantage of white light sensing is that ambient light does not affect its measurements. However, vibration does. So, they require a very stable platform. Currently, a number of medical parts manufacturers are using white light sensors to detect scratches on par ts. They are doing this by correlating white light measurements to "master results," which they generate by measuring near perfect parts with surface analyzers. Surface analysis capabilities will improve when white light "dot sizes" become smaller.

Wrist-mounted cameras: New wristmounted cameras for CMMs allow the system to capture data within 2.5° of normal on any part surface regardless of its orientation on the machine's table. In multi-sensor systems like this, users can employ tactile probes and iterative alignment algorithms accurately to locate parts within the measurement envelope before they measure any features with the camera. By using this approach, they can eliminate the need for a vision stage system and the need for expensive fixtures to locate measurement areas normal to a stage system's Z-axis.

Line laser scanners: Rapid line laser scanners provide a fast way to capture clouds of points for analysis. Although cross platform software will accommodate line laser sensors, they are not always appropriate for certain types of medical parts. On some systems, they require too much of an offset from the surface. Also, in measuring some types of features, the line scanner resolution is not fine enough to capture necessary characteristics.

Two Fast Paybacks

There are numerous everyday benefits associated with using vision and multisensor systems for the validation and control of medical parts and their manufacturing processes. Notably, the newest systems and software have two capabilities that are delivering substantial immediate paybacks for their users. They have to do with the parametric programming and the best fit analysis capabilities that have long been available with tactile probes.

Parametric programming eliminates measurement bottlenecks. Users have been responding enthusiastically to the increases that CAD-based, crossplatform systems offer in programming throughput. By working off-line on a CAD model, rather than on-line on an expensive machine, they streamline program development and free up valuable machine time for measuring.

The dual rotary table option for Hexagon Metrology's Optiv multi-sensor measurement system orients the part normal to the sensor for precise 3D data collection.

For manufacturers making families of parts, the ability to generate parametric programs yield exceptional results. A parametric program creates a set of rules that govern the generation of programs for every member of a part family. For each new member of the family, the programmer simply changes values in a parametric table and the software automatically creates the new part program.

In one instance, it was taking a programmer about 16 hours to create multi-sensor measurement programs for each member of a family of orthopedic parts. There were nearly 100 parts in the family. The manufacturer contracted Hexagon Metrology to write a parametric program for measuring the parts.

Creating the base program took the application engineer about 40 hours. With that done, the manufacturer can now automatically generate all of the other programs as needed. The cost of generating the parametric part program was minuscule in comparison to the overall time saved. Best of all, the company believes that it can now keep up with its production part validation and product development metrology requirements.

Best Fit Reduces Rejects

Because CAD-based 3D vision and multi-sensor measurement systems produce true 3D data, it is now possible to analyze the output they produce with advanced measurement and fitting algorithms. Of these algorithms, Best Fit is proving to be one of the most beneficial.

Best Fit mathematically analyzes bone screws in 3D to determine if they conform to the specification as a group.

For example, an orthopedic appliance may have six or eight holes for inserting bone screws. Rather than evaluating each of the holes individually, Best Fit mathematically analyzes them to determine if they conform to the specification as a group. Evaluating them as a whole rather than as individual entities can substantially reduce the number of rejected parts. This type of data can also provide actionable information that allows users to rework parts with greater precision, further reducing rejects and improving manufacturing profitability.

Using parametric programming and Best Fit analyses are two generalized strategies for getting a fast payback from the new vision and multi-sensor approaches. In addition, there are numerous application-specific paybacks obtainable by fine-tuning the metrology approach to use the best combination of equipment and sensors in combination with enterprise metrology software.

Applications Multiply

Across all industries, many vision and multi-sensor equipment manufacturers are exper iencing record sales. Even though the medical device manufacturers are under regulatory obligation to document and obtain approval for changes in their quality validation processes, sales are up substantially in this industry as well.

In the medical device industry a majority of the application development work for vision and multi-sensor systems is in the new products arena. Some of these new product applications include: filter membranes, orthopedic screws, articulating hip and knee joints, pipette systems, aspirators, inhalers and hypodermic needles, to name a few.

Substantial numbers of customers are ordering multi-sensor systems with dual stacked rotary tables to measure these sorts of parts. This configuration allows the software to orient the camera along the normal of almost any surface of the part making for accurate 3D measurement. Dual rotary tables also allow multi-sensor systems to measure most parts with a single setup to improve measurement accuracy as well as productivity. At least one manufacturer believes that rotary tables on multi-sensor equipment will be the norm before the year is out.

In addition, a number of medical device manufacturers have seen enough benefit in these new approaches that they have begun the arduous task of requalifying validation processes for existing parts to accommodate vision and multi-sensor measurement. As modern vision and multi-sensor part validation methodologies gain approval, the medical products manufacturing sector will become the growth trend leader for these rapidly evolving technologies.