Abstract
Source code accompanying the paper:
Klinik, M.A.A., Jansen, J.M. & Plasmeijer, R. (2019). Resource Scheduling with Computable Constraints for Maritime Command and Control. In MAST Asia 2019, Makuhari Messe, Chiba, Tokyo, Japan 17-19 June 2019. S.l.: MAST
DESCRIPTION
Scheduler for a variant of the Multi-Skill Resource-Constrained Project
Scheduling Problem (MSRCPSP). It uses a genetic algorithm to compute a set of
possible schedules for a given instance definition.
CONTENTS
- *.icl/.dcl: implementation of the scheduler and all utilities
- *.prj.default: rename these files to .prj to get project files for
compilation. The clean compiler modifies these files, and we don't want these
modification under version control.
- main.prj.default: project file for the web application
- mainIO.prj.default: project file for the command-line application
- test.prj.default: project file for the test cases
- simple-genetic-algorithm: genetic algorithm library, used by the scheduler.
- TestFramework: test framework library we use to run the test cases
- instances: some example scheduling problems, to demonstrate the scheduler
- More information on how to compile and run this program can be found in README.txt
SHORT SUMMARY
Maritime command and control currently trends towards reduced crew size with highly trained crew members, with the intention of allowing more flexible deployment.
As people no longer have fixed roles, dynamic scheduling of personnel and equipment is required.
In this paper we identify what kind of scheduling problem arises for this, and how to solve it.
We focus on the requirements of incident response and damage control scenarios, which can be anticipated but are impossible to plan out in advance.
We perform a literature study to put our scheduling problem in context and present an algorithm that satisfies the identified requirements.
One particular requirement we want to support is the assessment of schedules with user-defined, arbitrary computable quality metrics.
The estimation of how good a resource, be it a person or a machine, will utilize its capability for a given task should consider factors like weather, positions of resources, or equipment degradation.
This should go alongside classical metrics like length of a schedule's critical path.
Our work should be seen as a puzzle piece for the development of an integrated mission support environment.
