SIGAda Artificial Intelligence WG Meeting
Held at the ACM SIGAda
Summer Working Group meeting and
Washington Ada Symposium
June 27 1995
AIWG 1-day Workshop
The purpose of this workshop is to bring together software researchers and
practitioners from real-time, Ada and AI communities to exchange ideas,
results, and lessons learned, addressing the unique set of problems brought
about by the integra tion of Artificial Intelligence (AI) into conventional
embedded systems.
Artificial Intelligence technology is proving to be of value in solving complex
problems requiring reasoning on dynamic, uncertain, and incompletely specified
domains. AI has been successfully applied in a range of applications including,
for example, decision support systems and situation assessment - for example,
target classification and early warning systems - by allowing the manipulation
of vast amounts of information generated by modern sensors and intelligence
systems.
Although these next generation embedded systems possess some unique special
purpose requirements, they must be integrated with existing conventional
software per forming conventional real-time tasks. The integration of matured
AI methods and techniques with conventional software engineering remains
difficult and poses both implementation and conceptual problems.
Our ultimate goal is to understand what problems the embedded AI community
(specially the defense establishment) face and what solutions are potentially
or actually available to address these problems. As a result, we would be able
to identify potential technology that makes the practice of software
engineering more effective in solving embedded AI implementation problems. This
introduces new conceptual and engineering problems into traditional AI and
important software engineering challenges; these include:
- Systems design and intelligence integration
- AI problems and their solutions within the framework of embedded real-time systems.
- Integration of AI technology with conventional, complex, large software systems.
- Real-time AI computing
- Algorithms (e.g., performance within real-time constraints)
- Theory (e.g., deadline-driven reasoning with dynamically changing data)
- Software Engineering issues
- Problems in scaling up (e.g., from prototypes to full-scale deployed systems)
- Compliance with standard practice
- Adequacy of traditional AI languages and tools (specially CommonLisp)
- Adequacy of traditional software engineering languages and tools (specially Ada)
Submit a position paper (2-3 pages) on any of the subjects listed above (or
other related topic) by May 31, 1995. Authors will be notified of acceptance by
June 10, 1995 and will be given instructions for final preparation of their
papers/presentations at that time. (Submissions by e-mail using
PostSript will be accepted and encouraged)
For additional information and address to send submissions:
Dr. Jorge L. Díaz-Herrera
Software Engineering Institute
Carnegie Mellon University
4500 Fifth Avenue, Pittsburgh, PA 15213-3890
voice: (412) 268-7636
FAX: (412) 268-5758
e-mail: JDIAZ@SEI.CMU.EDU