Joining the dots

Point cloud processing – the ability to create an accurate 3D model from a collection of scanned data, has traditionally presented a number of difficulties from a design point of view. However, thanks to a number of factors, that may be changing, which could give industrial designers the freedom to explore new and endless variations of products. For more than a decade, engineers have been able to scan physical objects and create a 3D point cloud. A point cloud is exactly what the name suggests: a digital collection of points that represents a physical object. In the past, it's taken a fair amount of expertise to go from a point cloud to an accurate 3D model that is completely editable in a CAD program. But steady progress over the last 10 years, aided by somerecent breakthroughs, has made things a lot easier. According to Ping Fu, president and CEO of Geomagic, a leading provider of 3D software for creating digital models of physical objects: "Within automotive and aerospace companies, hospitals, consumer product manufacturers, design boutiques, electrical component makers, medical device providers and a host of traditional and newly minted entities popping up throughout the world, a quiet revolution is taking place. And it all starts with the point cloud." Until recently, the most tenuous part of the physical-digital loop was making the connection between 3D scan data and 3D CAD parametric models. Over the past two years, that connection has been simplified and streamlined by two major developments. One is the ability to capture and reproduce design intent for a physical object; the other is Parametric Exchange, which enables automatic native reconstruction of geometry, speeding product development time and increasing quality of the final model. 'Design-intent modelling' is fundamentally the ability to extract design intent automatically from a physical object. It extends reverse engineering from simply producing an accurate digital copy to extracting the original design intent from a scan of a physical model, kick-starting the process of modelling and adapting that design in CAD. Users can start with physical models and prototypes of a new design (such as a car body sculpted in clay) and quickly get to a digital, CAD-ready representation. The technology involved in design-intent modelling is fundamentally different from CAD. CAD focuses on creating and combining various geometric features. It is a prescriptive process, where everything is explicitly driven by the user and built by the system. Design-intent modelling searches for unknown information by structuring the scanned data from physical objects and extracting engineering features. It determines a valid structure from partial data sets, and compensates for measurement errors to generate perfect surface geometries. The ability to generate CAD-ready surfaces from scans of physical objects laid the groundwork for the other significant development mentioned earlier: parametric exchange. Parametric Exchange (PE) is a feature exclusive to Geomagic that enables parametric models created in Geomagic Studio to be transferred directly into 3D CAD software. Parametric Exchange completes the software bridge between point clouds and CAD. It provides an intelligent connection with CAD to enable automatic native reconstruction of geometry. With Parametric Exchange, parametric surfaces, datums and curves can be transferred from point-processing software to CAD without the need for intermediate neutral files such as IGES or STEP. Instead of duplicating CAD functionality, Parametric Exchange sends the parts of an assembly – in sketch, surface or solid form – to the CAD program in its native format, where it can then be edited, assembled and modified according to the user's specifications and downstream needs. There is no longer the need to fill in missing data lost in the transfer to a neutral format, or to reconstruct features and pieces of geometry from scratch. The Parametric Exchange process leverages the specific strengths of point-processing and CAD software. Point-processing software organises and processes point-cloud data to create CAD models, and CAD software enables users to further modify and prepare models for product design and manufacturing. So what are the potential implications for design engineers of this 'closed loop' between scanned data and 3D CAD? Theoretically, they are huge. Says Ping Fu of Geomagic: "The closed loop between scan data and CAD gives industrial designers the freedom to explore new and endless variations of products. Think of thousands of permutations of classic Nike shoes, individualized Mattel toys, or personalised fuel tanks for Harley-Davidson motorcycles… The ability to capture an existing design and quickly adapt it to new styles and purposes is critical to the evolution from mass manufacturing to mass customisation. The combination of design-intent modelling with parametric modelling provides many of the essential tools that will deliver on the promise of individualised design on an affordable, mass scale." The possibility exists, then, that this process could save manufacturers millions in tooling costs. Instead of recreating an expensive mould from scratch, for example, companies can scan an existing one, analyse the wear and tear, design an improved model, and manufacture new moulds in days instead of weeks or months. In addition, the ability to process point clouds into accurate digital models is also opening up new applications in 3D inspection, computer-aided analysis, and quality control. Companies now have the ability to almost instantly scan a product coming off an assembly line and compare it to the idealised CAD model to determine deviations and changes caused by tooling or moulds. Car companies can conduct finite-element analysis (FEA) and computational fluid dynamics (CFD) simulations based on the actual digital model of a physical as-built part or assembly, rather than the nominal CAD model. Maintenance, repair and overhaul (MRO) organisations can quickly assess damage to parts and fulfil the dual goals of better repairs and getting planes back in the air faster. In organisations such as Schneider-Electric and ITT/Goulds Pumps, point-cloud processing is driving major engineering changes. Schneider-Electric has progressed from dimensional analysis of singular parts to being able to analyse functionality for an entire assembly, saving countless hours and material across the company's 90,000-employee, 130-country operations. Meanwhile, at ITT/Goulds Pumps, point-cloud processing has generated quality control breakthroughs, including new processes to improve impeller balance and determine foundry process variation. The time it takes to analyse impeller balance has been reduced from hours to 10 minutes, and the ability to accurately measure wall thickness for moulds is delivering greater quality in much less time. Another major area that is benefiting from increased automation and accuracy in point processing is the medical field. The ability to transform scan data into models that can be manufactured rapidly is having a major impact in areas such as dental implants, orthodontics, prosthetics, hearing instruments, breast reconstruction, and cleft lip and palate treatment. Industrial processes based on point clouds have crossed the chasm into medical, where the ability to scan a body part or a dental mould enables doctors and dentists to tailor treatments to the exact physiological needs of their patients. The results include better fit and functionality of prosthetic, dental and hearing devices; less invasive and more personalised treatment; greater ability to analyse success of treatment and make adjustments; and better communication with patients through 3D imaging. Summing up, Ping Fu says: "The beauty of applications spawned from the point cloud is their seemingly infinite variety. A confluence of ideas and new applications is coming out of dialogue among industrial, medical and artistic communities." www.geomagic.com