In Brazil, a large number of companies have implemented the Last Planner system in order to improve the performance of their production systems. However, most research studies developed so far have emphasized the analysis of qualitative data, based on a small number of case studies. Therefore, there seems to be a good opportunity to learn more about production control by analyzing data from a larger sample of projects, based on performance measures such as the percentage of plans completed (PPC) and the causes for the non completion of work packages. This paper presents further developments of a research project that was first presented at the IGLC13 Conference. It is based on data from 133 projects. Some of them have been investigated in previous research projects at the Building Innovation Research Unit (NORIE) of the Federal University of Rio Grande do Sul (UFRGS). The remaining data have been provided by companies that have been using the Last Planner system for several years. At first, the average PPC and the reasons for the non completion of tasks were analysed using descriptive statistics. In a sample of 96 projects multivariate regression analysis was used for explaining the variance of weekly PPC, using as independent variables the causes for non completion of work packages, number of assignments completed and timing of the projects. The results indicated that a major problem in most projects is the lack of effective implementation of look-ahead planning. Moreover, for multivariate regression analysis, some causes for the non completion of work packages, such as work force and planning have a strong contribution for explaining the variance of weekly PPC.
Last Planner, performance measurement, multivariate regression analysis