Abstract:
This thesis reports the development of a master production planning system for a post-tension strand factory. This factory manufactures post-tension strands as Engineering-To-Order. The strands to reinforce the structure depend on the nature of each project such as load function, load bearing profile etc. In addition, each project may use a different engineer. As a result, the factory often receives the requirement of strands from each project rather late and too close to usage date. Therefore, production plans are often inefficient and unreliable. Data of past designs are collected and analyzed to formulate the relationship between the ratio of strand weight per area and engineering specifications that are known before the design process using one-way analysis of variance (ANOVA). With 95% confidential level, ANOVA suggests that the ratio is affected by four engineering specifications namely column span, depth of band beam, slab thickness, and super imposed dead loads. A linear regression model is then formulated to predict post-tension strand requirement in each project. In addition to the lack of demand projection, there are poor communication and unclear roles and responsibilities among related departments in the production of post-tension strands. To solve this problem, small group workshops were organized to develop standard work process an online master production planning system. The test of the online system during August to October 2012 resulted in 95% of the actual production that followed the master plan, compared to only about 60% in the past. Besides, the new planning system enables the planning of appropriate transportation modes.