A Review of the Twenty-Second SOAR Workshop

The proceedings typically are just copies of the slides used for the presentations, but sometimes more extensive prose or examples are included. In the last several years, these “slides” have been put up on the web.1 There were 37 talks this year as well as a discussion session with 57 attendees. The speakers came from 11 sites (including 2 from Britain), but the talks were not evenly distributed because 13 came from Soar Technology, 10 from Michigan, 4 from USC/ISI, and 2 from Pennsylvania State UniThe work presented at the SOAR workshop illustrated several trends that should be of interest to those outside the SOAR community and some of the issues that SOAR is grappling with that other groups developing unified theories of cognition or AI programming languages might find interesting as well. SOAR workshops are currently held annually in late May or early June. Recently they have been held at the University of Michigan and at the University of Southern California’s Information Sciences Institute (USC/ISI), two of the largest SOAR sites. They are held on a Saturday and Sunday, with a tutorial or two on the preceding Friday. This year the workshop was preceded by two days of tutorials: an introductory tutorial on Thursday and a more advanced tutorial on Friday. The tutorials were held at the University of Michigan’s Advanced Technology Lab and at the workshop site. The workshop was held just off campus at Soar Technology, a fouryear-old startup company that produces SOAR systems for industry and government. We think this is a good sign because it indicates a certain maturity of SOAR. This site also indicates that the largest SOAR site, as least as indicated by speakers at the workshop, is now a research and development company. The SOAR workshop assumes that the audience is already familiar with SOAR (thus the tutorials). The talks are short, sometimes as short as 5 minutes, rarely longer than 15 minutes. tems, (4) computer-generated forces, (5) games, (6) interfaces for programs, (7) interfaces for modelers and users, (8) reuse, (9) models of eyes and hands, and (10) robotics. This is our take on the workshop, and it reflects our biases and interests. The proceedings, based on a short discussion at the end of a session, will continue to include just slides, not full papers, providing a very good yearly summary but with some obvious defects. The format provides a lightweight way to describe ongoing work as well as projects that are about to start, which encourages collaboration.