An evolutionary optimization method to determine optimum degree of activity accelerating and overlapping in schedule compression

Compressing project schedule using activity accelerating and overlapping requires that an intensive time-cost trade-off analysis be carried out, to determine costs and benefits for each day of compression. However, the cost elements and implications of compression techniques differ significantly, since activity accelerating imposes extra direct cost whereas activity overlapping adds a risk of changes and rework. Such a trade-off becomes even more complicated in capital projects comprised of a large number of schedule activities and relationships. The variety of combinations of accelerating and overlapping of different activities in these complex networks can offer numerous possibilities for compression with various costs and potential risks. The lack of a reliable analytical tool for performing a precise cost-benefit analysis causes this critical task to be performed in a subjective manner during the planning stage of projects. The purpose of this paper is to present an advanced method using a multi-objective evolutionary optimization tool seeking the optimum degree of accelerating and overlapping during the schedule compression process. This optimization technique would be beneficial in maximizing project benefits while meeting the intended target dates. Key words: schedule compression, crashing, overlapping, substitution, accelerating, optimization, genetic algorithms. Raccourcir un echeancier de projet par l'acceleration et le chevauchement des activites demande une analyse circonstanciee comparative temps-couts approfondie afin de determiner les couts et les avantages pour chaque jour de compression. Toutefois, les elements de couts et les implications des techniques de compression different beaucoup, puisque l'acceleration des activites exige des couts directs supplementaires alors que le chevauchement des activites ajoute un risque de changements et de reprise du travail. Un tel compromis devient encore plus complique dans les projets d'immobilisations comportant un grand nombre d'activites programmees et de relations. La variete de combinaisons d'acceleration et de chevauchement des differentes activites dans ces reseaux complexes permet d'offrir plusieurs possibilites de compression presentant divers couts et risques potentiels. Le manque d'outils analytiques fiables pour realiser une analyse couts-avantages precise fait que cette tache doit etre realisee subjectivement durant l'etape de planification des projets. Le but de cet article est de presenter une methode evoluee utilisant un outil d'optimisation evolutionnaire a objectifs multiples recherchant le niveau optimal d'acceleration et de chevauchement durant le processus de compression de l'echeancier. Cette technique d'optimisation pourrait permettre de maximiser les avantages d'un projet tout en rencontrant les dates ciblees indiquees. [Traduit par la Redaction] Mots-cles: compression de l'echeancier, compression des delais, chevauchement, substitution, acceleration, optimisation, algorithmes genetiques.
Introduction Schedule compression is the technique by which project managers shorten project duration. Generally, schedule compression is performed using two main techniques. The first may be termed 'accelerating'. Accelerating is [...]