Decison support system for pallet utilisation in the sheet feeder industry

Abstract

This thesis aims at developing a decision support system for pallet utilisation in the sheet feeder industry. Furthermore, the research project has focused on providing an improvement to the current system of distribution of the customer order onto pallets. The overall research objectives are to: (i) Establish the state of the art in areas related to the project; (ii) Identify the key variables such as pallet sizes, truck sizes, material features and customers’ requirements for building the pallet utilisation model; (iii) Develop an analytical approach and a prototype system to improve the utilisation of the pallets and the truck volume; (iv) Validate the developed model and prototype system through real-life case studies and expert opinions. A combination of research methodology approaches has been employed in this research study. Firstly, a familiarisation stage has been conducted through comprehensive literature review and visits to the sponsoring company. The second phase was to conduct semi-structured interviews with the shop-floor operators and operations directors to have a better understanding of the business needs. This was combined with measurements of the trucks and the pallets and internal documentation from the company. In this stage, the researcher also distributed questionnaires to other sheet feeder companies to know if the system could be applied also to more companies. In the third phase, the data gathered were analysed for the subsequent development of the system. Then the system has been developed following the logic rules obtained by analysing the data. Finally, the developed model and system has been validated, through real-life case studies and expert opinions. The developed analytical model has been implemented in a prototype system. The developed system comprises a knowledge-based system, material handling modules, databases and user interface. The developed system for pallet utilisation encompasses pallet database, truck database and material features databases. The developed model and system has the capability to estimate the number of corrugated cardboard sheets that should be loaded per pallet. Additionally the system predicts the number of sheets per stack and the type of pallet that should be used. The system provides the shop-floor operators with an effective tool for pallet utilisation in order to reduce their decision time. Therefore, the developed system provides cost and time savings in terms of improvements of the pallet and vehicle volume utilisation for the preparation and delivery of the corrugated board product. This system has given to the sponsoring company an improvement of 5% in the way of orders’ management.