lerman@vaxc.ARPA (Lee Erman) (10/02/89)
Call for Participation WORKSHOP ON INTELLIGENT REAL-TIME PROBLEM SOLVING -- IRTPS SAN FRANCISCO BAY AREA NOV. 6/7, 1989 *** Technical Thrust *** AFOSR, RADC, and WRDC have begun an initiative to stimulate the development of a national basic research focus on Intelligent Real-Time Problem Solving (IRTPS). Despite large progress in knowledge-based systems, current technology does not meet the stringent requirements for predictable, high-quality results achieved in a timely manner in the presence of excessive demands for problem-solving resources. Recently, research has begun to extend knowledge-based systems technology to include real-time interactions with ongoing processes. This requires capabilities in several dimensions not addressed by current knowledge-based systems technology. These include - ensuring that the application will produce relevant output in a time appropriate for the environment, - continuous execution of processing facilities, - adapting the reasoning process and its guiding strategies to changes in the environment invalidate previous input or modify the available problem-solving resources, - graceful asynchronous interaction with the environnment, - interrupting ongoing reasoning processes and redirecting their attention to higher-priority environmental conditions, and - reasoning efficiently and effectively about temporal processes, including the reasoning process itself. In the first phase of this initiative, three groups (from Stanford, U.Mass., and Teleos) are working independently and together to help guide subsequent directions. These groups are studying IRTPS as a research domain and establishing preliminary ideas and answers to the questions in each of these areas: Area 1: Terms and Issues What technical questions should be included in IRTPS (and, conversely, which questions should be excluded from consideration)? What kinds of issues should be focused upon? Area 2: Common Experimental Problem: What would be a pragmatically viable and effective common testbed for the IRTPS research community? Area 3: Experimental Methodology: How should experiments be conducted and evaluated? *** Workshop Organization *** In this workshop, the three groups will present their preliminary results to a larger group of researchers in the IRTPS area, discuss these issues, and attempt to synthesize common answers to the questions in each area. The results will provide guidance for further directions by the initiating agencies; they will also be made available for public review. To accomplish this, the workshop will be organized as follows: I. Presentations and advocacy by each group of its proposed answers to the questions in all three areas. Each presentation is followed by comments and discussion led by a prepared, external reviewer. II. General discussion and preparation for area working meetings, including assignment of each workshop participant to one of the three areas. III. First area working sessions, in parallel. The sessions for each area will be led by a member of one of the groups. IV. First area summaries, in plenary session. V. Area refinement sessions, in parallel. VI. Final area presentations, in plenary session, and wrapup. The area participants may come to a single common opinion, which will be reflected by a single consistent statement of the answers to the questions in the area. Or there may be residual disagreement or alteratives. Shortly following the workshop, the area leaders will be responsible for submitting their reports, which the organizing committe will combine, edit, and present to the initiating agencies and for extensive public review. *** Logistics and Application *** The workshop will run from 9:30am on Nov 6. through 5:30pm on Nov. 7. To foster a workingshop, attendance will be limited. Because of the intensity and coherence of the workshop, applicants are requested to pledge their participation for the entire workshop. The workshop will be held at a location near Santa Cruz, CA. This is approximately a 50-minute drive from the San Jose airport and a 80-minute drive from the San Francisco airport. The workshop registration fee (which will be approximately $125) will cover the meeting rooms, two lunches, dinner on Nov. 6, and coffee breaks. Attendees will be responsible for their own lodging, other meals, transportation, and incidentals. > To apply for attendance at the workshop, each applicant (i.e., each individual) must submit a brief (2-5 pages) description of your IRTPS-related activities and interests. If possible, include your fax number and email address. You may submit by email, fax, or hard-copy. > If you are willing to be one of the group reviewers, which will require you to read the group's reports, present a review at the workshop and lead discussion, please also indicate that interest and willingness. Schedule: Oct 6: Deadline for receipt of applications by workshop chair. Oct 13: Notification of acceptance. Nov 6/7: Workshop (Note: The workshop falls on U.S. Election Day, Nov. 7. U.S. voters may wish to apply for an absentee ballot.) *** Organizing Committee *** Lee Erman -- IRTPS Chair Bill Baker, WRDC Cimflex Teknowledge Corp. Paul Cohen, U. Mass. Knowledge Systems Div. Michael Fehling, Teleos and Stanford 1810 Embarcadero Rd. Nort Fowler, RADC P.O. Box 10119 Yoav Shoham, Stanford Palo Alto, CA 94303 Abe Waksman, AFOSR 415/424-0500 X.422 fax: 415/493-2645 email: LErman@teknowledge.com -- Someday has arrived