From: Piermarco B. <pm...@cs...> - 2003-12-03 17:54:45
|
Hi, I am investigating the possibility to use OpenCyc for real-time applications, especially robotics. The question is: since Cyc is a huge knowledge base, even though I built my own "little" microtheory, would this suffise for getting good performances during query time? I mean, would this assure a reasonable computation time for inferencing any fact =66rom the KB when, say, the system is approaching a potentially dangerou= s state? Would Cyc let the robot get broken crashing against a wall, or would rather Cyc save it? Many thanks for any help. Best regards --=20 Piermarco Burrafato Ph.D. Student DINFO Universit=E0 degli Studi Viale delle Scienze 90128 Palermo, ITALY EMail bur...@cs... |
From: Ben Y. <by...@ya...> - 2003-12-03 18:17:28
|
I think the answer to this question might hinge on the garbage collector in Cyc's run time environment... it's certainly something that I'm curious about, in any case. How incremental is the GCing? If fairly incremental, what's a safe upper bound time wise on a collection cycle? Is it in its own thread? If so, what's the max fraction of cycles it would take away from the execution at any given point. Is there any way of configuring any of these behaviors? I'm curious about what you're doing robotics-wise, as I'm doing robotics research as well. Something to discuss one on one if you're so inclined. Regards, Ben --- Piermarco Burrafato <pm...@cs...> wrote: > Hi, > > I am investigating the possibility to use OpenCyc > for real-time > applications, especially robotics. The question is: > since Cyc is a huge > knowledge base, even though I built my own "little" > microtheory, would > this suffise for getting good performances during > query time? I mean, > would this assure a reasonable computation time for > inferencing any fact > from the KB when, say, the system is approaching a > potentially dangerous > state? Would Cyc let the robot get broken crashing > against a wall, or > would rather Cyc save it? > > Many thanks for any help. > > Best regards > -- > Piermarco Burrafato > Ph.D. Student > DINFO Università degli Studi > Viale delle Scienze > 90128 Palermo, ITALY > EMail bur...@cs... > > > > > > > ------------------------------------------------------- > This SF.net email is sponsored by: SF.net Giveback > Program. > Does SourceForge.net help you be more productive? > Does it > help you create better code? SHARE THE LOVE, and > help us help > YOU! Click Here: http://sourceforge.net/donate/ > _______________________________________________ > OpenCyc-devel mailing list > Ope...@li... > https://lists.sourceforge.net/lists/listinfo/opencyc-devel ===== From Leaf Sunlight and mist turn a young leaf into tea. Tea can turn you into something new. Tea. A natural gift of love. __________________________________ Do you Yahoo!? Free Pop-Up Blocker - Get it now http://companion.yahoo.com/ |
From: Stephen R. <re...@cy...> - 2003-12-03 19:55:49
|
The current release of OpenCyc would be suitable only for deliberative robot control activities, those in which some seconds are allocated for decision making. As a workaround, you could store a state-machine definition in a microtheory, extract it entirely into the robot application during initialization and use that for the sub-second reactive behavior of the robot. I am currently in the midst of writing a behavior engine for Cyc that takes this approach that may be available next year. -Steve On Wed, 3 Dec 2003, Piermarco Burrafato wrote: > Hi, > > I am investigating the possibility to use OpenCyc for real-time > applications, especially robotics. The question is: since Cyc is a huge > knowledge base, even though I built my own "little" microtheory, would > this suffise for getting good performances during query time? I mean, > would this assure a reasonable computation time for inferencing any fact > from the KB when, say, the system is approaching a potentially dangerous > state? Would Cyc let the robot get broken crashing against a wall, or > would rather Cyc save it? > > Many thanks for any help. > > Best regards > -- =========================================================== Stephen L. Reed phone: 512.342.4036 Cycorp, Suite 100 fax: 512.342.4040 3721 Executive Center Drive email: re...@cy... Austin, TX 78731 web: http://www.cyc.com download OpenCyc at http://www.opencyc.org =========================================================== |