jack@rml.UUCP (jack hagerty) (04/26/91)
In article <RSLAU.91Apr24132134@tarazed.usc.edu> rslau@usc.edu (Robert Lau) writes: >April issue of Discover, page 12... > >excerpt: > > PizzaBot was developed by engineers at Carnegie-Mellon to help disabled > people become entrepreneurs. [...] > > The result is a robotic arm placed in front of a semicircular counter, > with ingredients and tools within easy reach. After the disabled > manager calls out an order, a simple voice-recognition system decides > which of two possible sizes and 12 condiments it just heard. [...] > >Article said that K.G. Engelhardt was the head of the engineering team. > Yes, Engelhardt was the primary delveloper of the system. It was shown on TV on one of those "new technology" shows (Smithsonian or Scientific American). My only problem with this is that it is something of a grandstand stunt. Not that I have anything against rehabilitating the handicapped, but this particular system reduces the person to a sub-machine role. As you noted, the person does nothing more than speak into a voice recognition system which translates commands to the robot. You could just as easily have the customer place his own order on a touch screen which commands the robot directly. In fact, a major fast food chain is field testing just this sort of system. I found it fascinating that Round Table has parodied this system in their current commercial with a man-against-robot race. From a more entertaining direction, Nolan Bushnell (of Chuck E. Cheese fame) was working with Little Ceasers to develop a robotic pizza delivery system. I did the mechanical design on these little guys which turned out pretty well. They looked something like animated milk cans (for those of you in dairy country) and carried the pizza in their mid section. They rolled up to a "loading dock" near your table but you had to pull the pizza and other stuff (salads, drinks, etc.) out yourself. They also told jokes and did little routines. Most of the entertainment was done by a sub-species called "showbots" that had no food compartment but rather a hi-fidelity sound system in it's place. Even though the prototypes worked pretty well, both Nolan and Little Ceaser's had bigger fish to fry, so to speak, and the project died from inattention. - Jack ============================================================================= ||Jack Hagerty, Robotic Midwives, Ltd. jack@rml.UUCP (smart mailers)|| ||Livermore, CA ...!uunet!lll-winken!rml!jack (dumb mailers)|| ||(415) 455-1143 jack%rml@lll-winken.llnl.gov (desperate mailers)|| ||-------------------------------------------------------------------------|| || "The Biblical God is a sloppy manufacturer. He's not good at design, || || He's not good at execution. He'd be out of business if there was || || any competition." - Carl Sagan, _Contact_ || =============================================================================
hbg6@citek.mcdphx.mot.com (04/30/91)
>In article <RSLAU.91Apr24132134@tarazed.usc.edu> rslau@usc.edu (Robert Lau) writes: >>April issue of Discover, page 12... >> >>excerpt: >> >> PizzaBot was developed by engineers at Carnegie-Mellon to help disabled >> people become entrepreneurs. [...] >> >> The result is a robotic arm placed in front of a semicircular counter, >> with ingredients and tools within easy reach. After the disabled >> manager calls out an order, a simple voice-recognition system decides >> which of two possible sizes and 12 condiments it just heard. [...] I saw this system on some science television show, on the Discover channel I think. It was interesting but it got me wondering: Why do so many people/companies/universities try to throw robotic arms at every application, when a gantry style system would seem a better choice. I think this is the case with the pizza robot. The application of the sauce and cheese would have been simplified, the system would then be able to handle sliced pepperoni and other non-bulk toppings, and it would have cost a lot less. The point is not how to build a pizza, but I wonder why many engineers use expensive and complicated systems when simpler ones will work. For example, I have seen systems use an imbedded microcontroller, a driver module and a stepper motor in an application where a standard Bimba cylinder and an electronic valve would have worked fine. John ..................................................................... John Schuch - Motorola Inc., Computer Systems Division (602)438-3008 All opinions expressed are mine and not Motorolas, their loss. .....................................................................
nagle@well.sf.ca.us (John Nagle) (05/03/91)
>I saw this system on some science television show, on the Discover >channel I think. It was interesting but it got me wondering: >Why do so many people/companies/universities try to throw robotic >arms at every application, when a gantry style system would seem >a better choice. I think this is the case with the pizza robot. Gantry style robots have their own set of problems. I've worked with IBM's effort in that direction, the RS-1 (also known as the IBM 5640 Manufacturing System). Fast, precise linear actuators present problems. There's nothing as clean as Adept's electric direct drive system in the linear world. The RS-1 used multipiston walking hydraulic motors running on rails with wave-shaped teeth for each axis, along with ultrasonic magnostrictive delay line linear position transducers, connected via an analog servo system supervised by a computer, and powered by a hydraulic pump in a box a cubic meter on a side. It required 3 phase 220 volt power, compressed air, and chilled water. This for a machine with a working load of 15 pounds. It just took too much power to zoom that gantry around. Didn't sell. On a smaller, scale, Automatix's Robotworld uses 4" square units which translate in X and Y using two-dimensional linear stepping motors and hanging magnetically from a flat surface above the work area. The system is workable, but doesn't sell either. Neither of these systems offered force servoing or programmable compliance. For that you need a system with very little backlash and one that's back-drivable. There are good direct-drive electrical robots, such as the Adept SCARA-type arms and the little Zero arm. I don't know of a comparable machine in the gantry form. For food processing applications you need a machine that can be cleaned with boiling soapy water. Gantry machines tend to have lots of open track, complicated cabling and plumbing systems, and exposed lubricated surfaces. All these are unacceptable in a food processing operation. FDA-approved arms exist (the manufacturer is in Colorado Springs, but the name escapes me.) With arms, you can put the machinery inside and out of the way, which is a big plus in many applications. SCARA-type arms tend to be the machine of choice today where a gantry-type motion would be appropriate. They're available, simple, reliable, fast, and fully enclosed. Adept is the big US vendor. John Nagle
cphoenix@csli.Stanford.EDU (Chris Phoenix) (05/03/91)
In article <24564@well.sf.ca.us> nagle@well.sf.ca.us (John Nagle) writes: > On a smaller, scale, Automatix's Robotworld uses 4" square units >which translate in X and Y using two-dimensional linear stepping motors >and hanging magnetically from a flat surface above the work area. >The system is workable, but doesn't sell either. The Robotworld I saw required compressed air also. As I understand, the units are attracted to the surface by a magnet, and floated away from it by the compressed air for low friction. It also makes a lot of noise--the motors whine, and the air hisses. I never worked with it, but it looked pretty flexible, powerful, and precise. -- Chris Phoenix cphoenix@csli.stanford.edu #insert <funnyquote.h> #insert <graphic.h> #insert <stddisclaimer.h>