02. – 07. Juni 2002, Dagstuhl Seminar 02231
Scheduling in Computer and Manufacturing Systems
Auskunft zu diesem Dagstuhl Seminar erteilt
The objective of the seminar was to provide a forum for the discussion of current and proposed research in scheduling problems. It covered the entire spectrum from case studies of real applications to recent advances in mathematical foundations.
The seminar did not address only classical application areas such as distributed processing, operating systems, dependable systems, flexible manufacturing, etc. but also exciting new areas such as those in modern communications, examples being wireless networks, multimedia networks, and the internet.
The seminar proceeded along three broad fronts:
- applications, which includes empirical studies of existing systems as well as numerical studies of the analyses or simulations of system models;
- algorithmics, which includes the design and analysis of perhaps randomized algorithms ranging from simple and tractable on-line and greedy rules to methods based on semi-enumerative approaches, branch and bound, local neighborhood search, LP formulations, etc.;
- theory, which includes recent results in complexity classification, approximability, approximation schemes, analysis of classical problems under novel (or multiple) criteria, etc.
- 1. Applications .
- This topic includes both, empirical studies of existing systems in various application areas as well as numerical studies of the analyses and simulations of system models, and the study of the new problems arising in actual applications on new systems like cluster computing and grid computing. New characteristics like heterogeneity of the resources, large communication delays and hierarchy of scheduling have been investigated. In particular, the areas of application cover parallel and distributed systems.
This component deals with methods of analysis and modeling of distributed and parallel systems. Questions such as communication in MIMD systems, mapping program graphs and task systems onto various processor topologies, and load balancing are covered by the contributions. manufacturing and production systems.
This is the second broad area of applications where scheduling theory contributes important methods of modeling, analysis and algorithms. Shop problems in their most general form concern the manufacturing of a set of jobs on a set of machines where each job of the production process is characterized by its specific machine order. Contributions dealing e.g. with the optimization of production in flexible manufacturing systems and production centers with given numbers of parallel machines and as well optimization of robot control are welcome.
- 2. Algorithmics .
- This subtopic is mostly concerned with an algorithmic approach to scheduling problems. The unified framework for the presentations is the concept of computational complexity of combinatorial problems. The analysis of scheduling problems arising in computer systems and computer controlled manufacturing systems, and the adequacy of heuristic algorithms for solving these problems as well, have been discussed at the seminar. Additionally, methods employing artificial intelligence for solving some of the applications are covered by the seminar. This techniques are important to create a general tool for solving broad classes of practical problems.
- 3. Theory .
- This topic includes recent results in complexity classification, approximability, approximation schemes and heuristic approaches, and the analysis of classical problems under novel (and multiple) criteria.
Dagstuhl Seminar Series
- 04231: "Scheduling in Computer and Manufacturing Systems" (2004)
- 99431: "Scheduling in Computer and Manufacturing Systems" (1999)
- 9723: "Scheduling in Computer & Manufacturing Systems" (1997)
- 9520: "Scheduling in Computer & Manufacturing Systems" (1995)