Victor Scheinman: 
 Current automation practice is to produce high custom built installations with a system cannot be easily modified to meet changing demands. were adapted to advancements in technology. We need to design robot systems that are modular, portable and easily reconfigurable. to rapidly meet changes in technology for production requirements. Under the sponsorship of the US Air Force robotics and automation Center of Excellence race, advanced cybernetics group is developing tools for building robot systems quickly and cost effectively. The tools include a library of standard software modules. Each module has a defined plug and play interface to allow them to be easily replaced, to be modified and to take into account advancements in technology or changes in production requirements. The modules have been successfully employed by us as automation system integrators to rapidly generate robust robot systems. 

Col. Mike Leahy PhD: 
 The United States Air Force has a requirement to perform periodic maintenance on a large range of weapon systems in order to meet those requirements, an era of shrinking budgets and increasing internal and external competition. We manufacturing processes must be upgraded. Unfortunately, most of our depo remanufacturing processes are not suitable the classic robotic and automation solutions. Take a couple of seconds to talk to you about the different constraints that may actually make that happen. First of all, we deal on small batch sizes. We don't have 1000s of cars and millions of widgets, we have 10s of planes to work on a particular year and the devil process. We have a lot of feature uncertainty both in the placement of where the part or workspace in this case the plane is going to be when we work on it and what we're going to be working on in the plan itself. We don't have CAD models to help us with that. We're gonna have to do this in an era having a existing workforce still in place, and I want to do it with high reliability maintainability custom solutions in the past didn't work you got to be able to call up and get a response get something fixed. And all this has to be done with low insertion costs. 
 
 The Air Force Materiel Command robotics and automation Center of Excellence at Kelley Air Force Base in San Antonio, Texas, that unified tele robotic architecture program for Utah. under the aegis of Utah advanced cybernetics group ACG was contracted to validate our fundamental thesis, which is that existing commercial systems could be cost effective, they extended meet our existing telerobotics requirements and provide a base for future evolution, a common set of core telerobotics capabilities. I talked about x toolkit if you will, emerges at the top of that list is the ability to fly above a surface and more technical terms autonomous regulation. of separation of standoff distance, in the direction perpendicular to a work surface, while the operator controls the motion parallel to the surface maybe with a joystick maybe not. Second is the ability to follow the surface. Again in techno speak, autonomous regulation of force and the direction perpendicular the work surface while the operator controls motion parallel to the surface that allow you to move along and do some surface finishing visual registration on the workpiece. Quite simply, I happen awareness I want to be able to work on maintaining an angle of attack a lot of times the tool has to be held at a certain angle I got to maintain that angle between the tool and the workpiece upon this regulation of angle between the tooling axis and the workpiece normal again while the operator controls the trajectory. Finally, the ability to learn workpiece dimensions by visual based tracking of a workpiece trajectory might want to just go out and mark the area that I want to cut out but a black grease pencil want to be able to follow that with a vision system it'll memorize that locations that are part and be able to bring it back in and be able to work on it later. For the purposes of our demo we just did dolls in a planar system. 

Victor Scheinman: 
 The ACG efforts you have witnessed in this tape have not been done in a vacuum. In fact these efforts are connected to a larger overall UTAP program, which have witnesses the output of a phase two demonstration phase, the purpose of which was to initially validate the gap on this you tap architecture. That architecture has been under development for about the last year and a half. And it's closely linked to other JPL on this architecture reference, such as the enhanced machine control or EMC project. It's also linked with other national efforts such as that sponsored by the National Center for manufacturing, sciences and CMS. They are currently running the NGC program and under that umbrella our tele robotics architecture can be looked at as an application architecture. 
 
 In my many years of experience in the field of robotics, I've seen numerous demonstrations and countless papers on the subject of general purpose robotic automation. Few of these embryonic thoughts and examples filtered down to solid factory or hangar floor applications because of the myriad of reasons mostly relating to the failure to effectively deal with all of the real world applications issues. The unified teller robotic approach and the tele robotics toolkit technology will close the gap between demonstrations of concept and practice. In fact, we already have some concrete examples of practical applications. at Warner Robins Air Logistics Center robotic CO2 paint stripping has been tested on F18 wings using surface sensing and motion planning tools. 
 
 At IBM hard disk platters are rapidly and accurately loaded into sputtering fixtures using sensing measuring and surface mapping tools from the toolbox. 

John Bianchini IBM:
 IBM ACG, gently develop the software system for multiple loads of disk drive platters into the old back spreading machine. IBM require the system to be agile. That is the software has to be written so that it can be inexpensively and rapidly reconfigured for different sizes and palette geometries. The software was initially developed by millimeter blind. The initial changeover to 95 millimeter discs was accomplished in two days. 

Victor Scheinman:
 At the Schlage Lock division of Ingersoll Rand Corporation, rough cast door levers are being sanded and polished with the robotic system using force sensing and surface contouring tools involved as part of this program. UTAP is an excellent example of how government and industry can cooperate to produce key enabling technologies that will provide immediate utility value to the government and return the United States. robotics industry to world class prominence.