The achievements in robotics and artificial intelligence we have seen in recent times are inspiring. However, scientists remain at a very early stage in our attempt to replicate the human visual system in machines. As a fellow in the Mind-Brain Institute at Johns Hopkins Medical School, I have been specializing in the neuroscience of vision, lately with the help of 3D digital scanning and the Optix hardware and software we acquired from 3D Digital Corp.
At a point in our research into this subject, a colleague and I began researching 3D laser scanning with great anticipation at what it could provide. Neither of us had experience or background in 3D scanning, so we took some time to study the market and the various vendors. We came upon options that seemed to provide what we needed, but at a purchase price that was far beyond our budget. We also looked at 3D scanners that were affordable, but not adequate for our needs. Finally my lab manager and I took a look at 3D Digital Corp., and were encouraged by the combination of price and technical sophistication in both the eScan and Optix units. We visited the company headquarters for a demonstration and learned a good deal. Soon after, there was a more formal presentation and training session at our lab in Baltimore, using the Optix model we selected.
For reasons of non-disclosure that are routine prior to completion of experiments and publication of the results, it isn’t possible to share many details of the research work we’ve done using 3D scanning. In general, it involves the elements of visual neuroscience that allow shape recognition and object perception in the higher visual centers of the brain. The application of advanced technology that could be developed from work like ours is, in the best-case scenario, dramatic and valuable. It involves computers and robots attaining much more human-like capacity for recognition and response to objects, for example in micro-surgery. The field is still at an incipient stage—it is agreed that large rooms filled with mainframe computers still cannot match the acuity and capacity found in the brain of a normal three-year-old human.
Our first field work took us to the Walters Art Museum, to scan an abstract sculpture. We performed the scan and imported our file using the SLIM software. The museum was very cooperative, but it was made clear to us that a limited time slot would be available. We had to capture and import our files in one attempt, which we were very pleased to be able to do. The consults and training we received from 3D Digital engineers made it possible for us to achieve our goal in one shot.
(Aug. 3, 2011) Well-engineered parts and components are vital for any machine assembly and maintenance. In the world of rotary-wing aircraft, this is of absolute highest priority. Our company is currently working on special maintenance projects involving the Agusta Bell 2065 and the Boeing CH-47 series of helicopters. Each of these aircraft was developed in the mid-20th century and has proved both valuable and versatile. That is why they are both still around and needing replacement parts.
(June 6, 2011) The world needs clean energy like wind power, and the wind-energy industry needs affordable, top-quality wind turbine blades. The Department of Industrial and Manufacturing Systems Engineering at Iowa State University is using a 3D Digital Optix 400L laser scanner to help solve that need. When you see wind turbines high above Midwestern fields, it’s a pretty sight. But the reason you can see them from such a distance is their mammoth size—one blade is about 40 yards long and weighs over 10,000 pounds. One viable material to build them with is fiberglass, but it’s labor-intensive to make these blades, so manufacturing has traditionally been done in low-cost venues outside the U.S—but then they’ve got to be shipped.
I was introduced to 3D laser scanning at Penn State, where my PhD dissertation advisor introduced me to reverse engineering and rapid prototyping using this new technology. Early use at Iowa State for the Optix scanner was in a project for the rapid manufacturing of service parts for agricultural equipment; research in a subtractive rapid prototyping process called “CNC-RP”. The idea was to be able to deliver fully functional service parts for perhaps decades-old equipment that is still in use (a challenge for both agricultural and military equipment, in particular). This is necessary when the parts are unavailable, would require long waiting periods, or would require extensive inventories in warehouses — perhaps not acceptable when a crop needs harvesting right away or costs are prohibitive. In one example, we created a linkage for a harvester, scanning the failed component, importing it to a CAD program, modifying the file to “fix” the broken section, then rapid machining the actual new part in-house. From arrival of the failed part to boxing and shipment of the new steel part was 48 hours total.
