no title

Subject: no title
From: Juergen Dorn (
Date: Fri Jun 12 1998 - 19:54:15 MET DST

Call for Papers

Special Issue on Machine Learning for Scheduling Applications

The European Journal on Artificial Intelligence

Modern production techniques as well as the required production management
has changed in the last decades from mass production to a specialized
production in small batches with a vast amount of various products. To be
competitive an enterprise must react immediately on market demands and must
guarantee high quality standards. The global competition demands also an
improved efficiency in the production process which is tried to achieved by
avoiding intermediate stocks and waste production (both, total quality
management (TQM) and business (re-)engineering (BPR) put emphasis on these
goals). Moreover, modern production management asks for controlling also
the whole supplier chain.
   The trend in business information technologies is to apply standard
software (e.g. SAP/R3). However, shop-floor scheduling was to our
knowledge never been realized in such a software package. This seems to be
due to the diversity in products, processes (manufacturing technologies),
constraints and objectives. A further obstacle are the regular
modifications of products and processes. Current practice is to apply a
kind of high-level languages such as mathematical programming or constraint
languages to solve industrial scheduling problems. Usually, these programs
are outdated soon.
   One technique to broaden the applicability of systems and to adapt them
continuosly to the new environment could be machine learning. First
academic systems were already described that learn heuristics for special
applications. We are looking for papers addressing various types of
knowledge learnable for production process scheduling. Papers may address
- learning and adapting of heuristics
- learning of constraints and its weights in multi-criteria scheduling
- learning of new processes and routings
- learning of criticality of orders
- recognition of bottleneck resources
- etc.

Authors are invited to submit a full paper (up to 15 pages).
A Latex-style file and formatting instructions can be obtained from:
Electronic submissions in postscript format sent to the guest editor are
strongly encouraged. Alternatively, five hard-copies should bereceived
before September 1, 1998 by the guest editor.
Submissions are refereed according to the AI Communications refereeing
policy. Author will be notified of acceptance of their paper by November
13, 1998.

Guest Editor
Juergen Dorn
Technische Universitaet Wien
Institut fuer Informationssysteme
Paniglgasse 16
A-1040 Wien, Austria
Tel: +43 (1) 58801 6123
Fax: +43 (1) 5055304

Important Dates
Submission deadline September 1, 1998
Acceptance decision November 13, 1998
Final manuscript December 18, 1998
Publication March, 1999

This archive was generated by hypermail 2b25 : Fri Mar 03 2000 - 16:17:15 MET