Date on Master's Thesis/Doctoral Dissertation

5-2009

Document Type

Master's Thesis

Degree Name

M. Eng.

Department

Industrial Engineering

Committee Chair

Depuy, Gail W.

Subject

Industrial management--Data processing

Abstract

A common problem faced by most organizations in today's world is one of worker-task assignments. Assigning a large number of complex tasks to workers at various training levels can be a complicated process which has the potential to cost or to save a company large sums of money. The aim of this project is to develop a heuristic tool designed to match tasks to workers given the workers skills proficiency profiles. This heuristic should also provide a training plan which will rectify current worker skills gaps while minimizing training costs. Prior research maintained a focus on utilizing mathematical models of this skills management problem. The main difficulty with these mathematical models is that they were unable to reach feasible solutions in a reasonable amount of time when the problem size became large. It is therefore wise to investigate possible heuristic solution techniques. This research will compare and contrast three specific heuristic techniques: a Greedy Assignment Algorithm, Meta-RaPS Greedy Heuristic, and Meta-RaPS Shortest Augmenting Path (SAP) Heuristic. Meta-RaPS is a meta-heuristic that is used to improve the performance of algorithms by strategically infusing randomness which allows the exploration of more of the solution space. The skills management heuristics developed in this research were tested using 47 randomly generated data sets generating results within 0.03% of optimal for the recommended Meta-RaPS SAP solution methodology.

Share

COinS