The Executive Committee of the Operations Research Society of South Africa awarded Berndt Linder first place in the competition for best oral presentation by a masters student at the 42nd Annual Conference of the Society, which was held at the Protea Hotel Technopark in Stellenbosch. The prize is a one year membership of ORSSA for the calendar year 2014. Berndt Linder is part of the Unit for Systems Modeling and Analysis (USMA) research group. The thesis topic and abstract is presented below.
Thesis research topic:
A Model to Optimise the Linked Sawing and Ripping Decisions in the South African Pine Wood Industry.
This paper looks at how to model and link two important cutting operations in the South African pine wood industry. The model is proposed to help answer the question: given the current log supply – what should some of the decisions, specifically machine settings, be for the two cutting centres to try and meet market demand at a minimal cost. The first decision is how to saw up the log supply into its different thicknesses by choosing certain cutting patterns; the second is to decide on a rip saw’s settings (priority values) which determines how the products from operation one is cut into products of a certain demanded thickness and width. The objective function to minimise includes the raw material waste cost, and an over/under production cost. The over production cost is estimated to represent the stock keeping costs. The under production cost is estimated as the buy-in cost of purchasing the raw material from another wood supplier. The modelling tools used in this paper are the Probability Based Incremental Learning (PBIL) algorithm and Mixed Integer Programming (MIP) and the modelling software used are Simsaw and AIMMS. The model performs well against current decision software in South Africa, namely the Samwmill Production Planning System (SPPS) package, which combines simulation (Simsaw) and Mixed Integer Programming techniques to maximise profit. The model adds further value in modelling The model adds further value in modelling and determining the ripping priority settings linked to the decided upon cutting patterns.
Congratulations Berndt Linder!