There’s a quiet revolution taking place in design, engineering and manufacturing. It’s happening behind the scenes at leading automotive, aerospace and commercial product companies; at dental and hearing device companies; and within product inspection departments of industries ranging from turbine to electrical manufacturers.

It’s disruptive in its impact on speed, product differentiation and quality, while being complementary to established CAD/CAM/CAE processes. Everything it touches tends to get better by association. It’s called digital shape sampling and processing (DSSP), and it’s driving a convergence that is forever changing the way we design, analyze, manufacture, inspect and maintain products.

In a recent keynote speech at Geomagic’s worldwide user conference, Joel Orr, vice president and chief visionary of Cyon Research Corp., characterized DSSP as the convergence between “simulated and real, shape and number, intended and actual, past and possible.”

What is DSSP?

DSSP is a category name that encompasses the convergence of multiple technology advances. It describes the ability to use scanning hardware and processing software to digitally capture physical objects and automatically create accurate 3D models with associated structural properties for design, engineering, inspection and custom manufacturing. What digital signal processing (DSP) is to audio, DSSP is to 3D geometry.

DSSP has evolved as a result of several technology areas that have matured over the last decade, including optical 3D scanning, reverse engineering, computer-aided inspection, and geometry processing. The demands within these areas have led naturally to integrated DSSP solution offerings.

As with bioscience and other fields with large market potential, DSSP has attracted some of the world’s leading scientists and major research funding from government agencies. Geomagic, for example, has Dr. Herbert Edelsbrunner and Dr. Tamas Varady, with more than 200 published papers between them, actively involved in research and development. The company has received millions of dollars in funding from the National Science Foundation and the National Institute of Standards and Technology to further develop key DSSP technologies and apply them to industry.

The essential components

DSSP requires two essential components: scanner hardware to capture point data, and software to process point data into useful digital results.

Technology advances made by manufacturers of optical scanners during the last decade were the first steps in making DSSP possible. Previously, engineers were limited to manually capturing one point at a time. Optical scanners have made it possible to collect millions of points in the time it used to take to record a few points. DSSP enables capturing the entire bounding surface geometry of a physical object – including product features, colors and even textures.

Mechanical data collection methods such as those used by coordinate measurement machines (CMMs) are still important to some inspection applications, but increasingly all forms of inspection are moving from contact mechanical to non-contact optical technology.

Gathering millions of points of data has little or no value, of course, unless the data can be processed easily into digital models with the quality needed for downstream applications. That’s where software plays a critical role.

The combination of greater price/performance for desktop computers and innovation in geometry processing algorithms has moved DSSP forward at a breathtaking pace. Point-cloud data that would choke a high-end computing system five years ago is now easily digested by modern PCs. Gaps and noise in scanning data that used to take days to resolve are now corrected automatically in the best DSSP software. Conversion to polygons and NURBS surfaces, once requiring days of tedious work, can now be handled in minutes using a natural, intuitive workflow. Interaction between parametric CAD software and programs such as Geomagic Studio and Qualify is fast and intuitive.

Accurate repeatability of DSSP software is making it possible, especially in applications such as digital quality inspection, to move analysis and reporting tasks from experts in offsite offices to staff on shop floors. Experts can now spend more time on product development and manufacturing design processes. Automated reporting using 3D graphics in standard formats enables inspection results to be easily understood and shared throughout the enterprise.

A complement to CAD/CAM

DSSP has faced misunderstandings in relation to CAD/CAM. Far from being an overlapping or competing discipline, DSSP complements CAD/CAM and is an essential part of the digital design and manufacturing life cycle.

With its roots in drawing, CAD/CAM software is limited to prescriptive modeling methods. In other words, pre-defined geometry must be prescribed by an expert to a software tool for the purpose of modeling. CAD/CAM starts in the virtual world with a goal to produce better products in the real world.

As a drawing-based technology, CAD starts with a blank screen, requiring that the user input dimensions, shapes, curves and surfaces that will define an object. It is great for modeling new products, particularly those with simple facets and standard geometric shapes. It is limited, however, when it is faced with describing or representing the complexity of the existing world.

With its roots in imaging, DSSP offers descriptive modeling methods. The software extracts geometry and topology from measurement data and describes them to users for archiving and reuse for multiple purposes. DSSP starts in the real world with a goal to produce high-quality digital models in the virtual world that can be used by CAD/CAM/CAE applications.

DSSP bridges the gap between the point domain of measurement and the shape domain of design. It aligns the physical and digital worlds, ensuring that the design model is an accurate representation of the as-built product. This alignment is often missing in CAD/CAM, where changes needed to adapt a design for manufacturing create differences between the CAD model and the physical product. DSSP closes the physical-digital loop.

Accurate alignment between the digital representation and as-built product delivers major benefits, including the following:

• Time and money savings due to faster development cycles and design iterations.

• Better quality through more accurate engineering analysis, resulting in less manufacturing waste, lower rework costs, and reduced product returns and recalls.

• The ability to customize products in mass quantities, creating competitive advantage and distinctive product branding.

• Improved tooling and product design based on the ability to capture and document the wear and tear of everyday use.

• Automated quality inspections that reduce labor costs and staffing requirements.

• Replacement of tedious, manual jobs with automated processes, providing more control, less reliance on outsourcing, and greater employee safety and satisfaction.

High-profile applications

The role of DSSP in NASA’s return-to-flight initiative is an example of how this technology has a profound effect on the way we capture, process and use 3D shape information. Beginning in August 2005 and with each successive shuttle mission, Geomagic DSSP software has given NASA the ability to detect, assess, repair and validate a repair in the unpredictable environment of space.

As the shuttle nears the space station on the second day of flight, it rolls over to expose its underside. An optical scanner attached to a 50-foot-long extension of the shuttle’s robot arm scans the underside of the shuttle’s wings to capture damage. Scan data is transmitted to Houston, where Geomagic software is used to create 3D models of the damaged tiles from the data. The models are then analyzed to determine the extent of damage. Fortunately, damage in shuttle missions to date has not been consequential enough to require actual repair.

If damage is ever considered too extensive for safe reentry, the situation could call for a spacewalk by astronauts to make the repair. In this case, the test tiles would be used to develop a step-by-step repair process.

During her keynote speech at Convergence 2007, Col. Eileen Collins, retired commander of the space shuttle Discovery, recognized the importance of DSSP in eliminating what used to be a major game of chance for astronauts returning from space. “Thank you for making space travel safer,” she told the assembled Geomagic developers, users and partners. “Not only did you do it, you did it fast.”

Another dramatic example of DSSP’s impact is a recent project to digitally recreate the Statue of Liberty. After the World Trade Center attacks, the value and vulnerability of U.S. national monuments received new consideration. Although there are numerous photos of the Statue of Liberty, there are no detailed architectural drawings that would enable an exact replica of the monument – especially details such as its artisan-crafted robe and realistic skin – to be re-created accurately.

Texas Tech University, in cooperation with the National Park Service and the Historic American Buildings Survey, used DSSP to capture the statue’s unique architecture. Texas Tech collected the data using a large-format 3D optical scanner capable of capturing 800 points per second and tested at 6mm accuracy. Researchers spent four 14-hour days scanning different points around the statue. They then took the scanned data back to Texas and ordered a computer with dual 1.8-GHz processors, a 3-GB RDRAM video card, and two 80-GB SCSI hard drives to process and register the immense point-cloud data set.

The Texas Tech team had 13 different scans with approximately 1.23 million data points per scan, a total of 16 million data points. Geomagic software enabled the university’s researchers to register and align the 13 scans, create polygon meshes, and generate the NURBS surface models that could be imported into a CAD program.

According to the project director, trying to register the points without the specialized software would have added at least 120 additional hours of work, and Texas Tech would not have been able to achieve adequate accuracy for the model.