By Topic

Advanced Software Technologies for Scheduling, IEE Colloquium on

Date 26 Apr 1993

Filter Results

Displaying Results 1 - 8 of 8
  • The future of scheduling-DAI?

    Publication Year: 1993 , Page(s): 8/1 - 8/2
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (124 KB)  

    It appears that AI scheduling research is moving outwards in a number of different directions: constraint-based reasoning, distributed AI, stochastic search, iterative repair, look-ahead techniques. The author is not suggesting that the community is `pulling apart', merely that the net is being thrown further, and to good effect. He believes that there is another promising direction to be explored. When we solve some combinatorial problem, the order that we choose to make our decisions may have a profound effect on the difficulty in solving that problem. This `order' is sometimes referred to as the instantiation order. Clearwater, B.A. Huberman, and T. Hogg (1991) published a paper on a distributed solution to the constraint satisfaction problem. We are given a single problem, and a number of agents, where each agent is capable of solving that problem on its own. The agents are each given a copy of the problem but attempt to solve it using different instantiation orders. That means that the agents navigate through the same problem space, but start from different positions. During the process of exploration the agents are allowed to share their discoveries. The authors claim that this can result in a super-linear speed up, or a `combinatorial implosion' View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Uncertainty in scheduling: probability, problem reduction, abstractions and the user

    Publication Year: 1993 , Page(s): 7/1 - 7/4
    Cited by:  Papers (3)
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (260 KB)  

    In most realistic scheduling situations, the information available to the decision-maker is both incomplete and uncertain. This complicates the automation of intelligent reasoning systems in the real world. The author discusses the issues involved in reasoning in uncertain environments and argues that scheduling is essentially a problem of decision-making under uncertainty. She classifies various types of uncertainty and proposes techniques to address these problems within the advanced software domain. These techniques include the use of probabilistic modelling, problem reduction, temporal abstractions and the user. Relevant issues are illustrated through an advanced scheduling system being developed at the University of Geneva. The author also mentions how these techniques relate to TOSCA, a system developed for manufacturing by AIAI View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Object oriented constraint programming for transportation problems

    Publication Year: 1993 , Page(s): 4/1 - 413
    Cited by:  Patents (22)
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (484 KB)  

    Constraint satisfaction techniques are a very powerful way to tackle highly combinatorial problems such as vehicle routing or crew scheduling. The author presents PECOS, an object oriented constraint programming library, used in various industrial applications, including transportation. It is available in two versions: a C++ library, or a Lisp library. PECOS offers some unique aspects such as: object oriented programming; expressing a constraint at the class level, which is shared by all the instances of the class; use of global optimization techniques such as simulated annealing; symbolic constraints for resource allocation problems; and so on. These features are described and illustrated by the CARAIBE system, which deals with locomotive scheduling for a French railway company. The inputs are the departure and arrival time and stations of all French trains. The system minimizes the number of locomotives and produces a detailed planning for each locomotive. The use of object oriented programming with class constraints allows the representation of the problem with five constraints only, whatever number of trains are considered. The use of simulated annealing enables a very good optimization. This application has clear counterparts within the air transport companies View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Classification of scheduling problems and selection of corresponding constraint-based techniques

    Publication Year: 1993 , Page(s): 1/1 - 1/3
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (204 KB)  

    It has been argued that the use of constraint-based techniques and tools allows the implementation of precise, flexible, efficient and extensible scheduling systems: precise and flexible as the system can take into account any constraint expressible in the constraint language; efficient in as much as highly optimized constraint propagation procedures are now available; extensible as the consideration of a new type of constraint may require (especially in an object-oriented framework) only an extension to the constraint propagation system, or the implementation of additional decision-making modules (without needs for modification of the existing code). Yet there are a number of tradeoffs (between precision, flexibility, efficiency and extensibility) to consider when designing a scheduling system. These tradeoffs relate to: the representation of the scheduling problem; the set of control strategies to make available for the resolution of the problem; and the relation between the scheduling system and other information providing systems operating in the environment. Each of these three points suggests a different classification of scheduling problems and association of constraint-based techniques to scheduling problem classes View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Two problems-two solutions: one system-ECLIPSE

    Publication Year: 1993 , Page(s): 3/1 - 3/3
    Cited by:  Papers (1)
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (180 KB)  

    The constraint logic programming system ECLiPSe is discussed. Computation in ECLiPSe alternates between two modes: constraint handling and host program execution. The host programming language is (an extended) Prolog, which handles search, and interaction with the programming environment. The control for host program execution mode is the usual Prolog control. The constraint handling mode has a quite different form of control, which generalises data-driven computation. During constraint handling all possible information is extracted from the constraints. When there is no more information to be extracted, the system returns to host program execution, which continues until another constraint is posted and constraint handling restarts. The authors show the advantage of combining Prolog programming with constraints handling for a shift planning application. They also indicate how to control the constraint handling itself, and the application of such control for optimising job-shop scheduling programs View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • RAMSES: decision-aid for the security of the 1992 Winter Olympic Games

    Publication Year: 1993 , Page(s): 6/1 - 610
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (384 KB)  

    Over a million spectators descended on Albertville in France for the Winter Olympic Games in February 1992. Two billion more watched the Games on television. France was in the limelight, and everything had to run smoothly. The venue made security and safety very difficult. In the event of any calamity or accident, the security forces would have to be rapidly redeployed from their normal duties. If several incidents were to occur together, the situation would be unmanageable by conventional means, and a system to co-ordinate operations would be needed. RAMSES was this system. RAMSES is a new type of computer application, integrating expert systems, a GIS database and a resource allocation system on distributed hardware. Conceived for the management of security at large events, it could readily be applied to other circumstances in which the activities of separate teams with disparate skills must be coordinated. An example is the handling of environmental disasters. Yet RAMSES is a decision-aid and not a command and control system. Its task is to collate, interpret and display information, offering the human decision-maker the best expertise in evaluating it and devising his plan but leaving him firmly in control View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Using heuristics for constraint satisfaction problems in scheduling

    Publication Year: 1993 , Page(s): 2/1 - 2/3
    Cited by:  Patents (1)
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (188 KB)  

    Scheduling problems such as job shop scheduling and timetabling can be expressed as constraint satisfaction problems (CSPs) and so are, at least in theory, amenable to solution by standard constraint satisfaction algorithms. Unfortunately, CSPs (and most scheduling problems) are in general NP-complete. Hence we should expect the time required to solve a CSP to increase exponentially with the size of the problem, so that in practice there will be a limit to the size of problem which can be solved in a reasonable time. One approach to tackling a problem which is too large to be solved exactly by a standard search algorithm is to devise algorithms which abandon the idea of a systematic search of the entire search space. Typically these algorithms form an initial solution which is deficient in some way (infeasible or suboptimal), and attempt to improve it by making local changes or repairs. The algorithm moves through a sequence of solutions, each of which is a neighbour, in some sense, of the preceding one. Several papers have described repair-based heuristic methods of this kind for constraint satisfaction and related problems. The author describes a heuristic of the same type, originally developed as part of the ROSA system (B.M. Smith, S. Bennett, 1992), which produces weekly anaesthetists' rotas for a hospital anaesthetics department View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • IEE Colloquium on `Advanced Software Technologies for Scheduling' (Digest No.163)

    Publication Year: 1993
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (268 KB)  

    The following topics were dealt with: classification of scheduling problems and constraint-based techniques; heuristics for constraint satisfaction problems; ECLIPSE, constraint logic programming system; object-oriented constraint programming for transportation problems; RAMSES decision-aid for security; uncertainty in scheduling; and the future of scheduling and AI View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.