Improvement model applying Time Series, TPM and SMED to reduce waste in a timber company in the Peruvian jungle

The main objective of this study is to analyze the problem of shrinkage generation in the lumber production process, impacting the sawmill’s costs and efficiency. The company under study showed a 9.55% of shrinkage generation, while the standard is 7%, which leads to a monetary loss of ${\$}$ 96,383: Raw material expiration (67.3%) and lost production time (27.5%) and other non-controllable factors (5.2%). To reduce the impact of the problem, we used Demand Forecasting, SMED and Preventive TPM techniques, which resulted in a reduction of the shrinkage percentage by 15%. Demand forecasting helps us to study the behavior of monthly sales, to discover the time series model that follows the data month by month, for the case study, simple and triple exponential smoothing. In this way we estimate the future, helping us to reduce the shrinkage due to overstocking as we would process reasonable quantities plus a safety stock to cover the uncertainty. Also, for the species that annually produce almost constant quantities, it was decided to use the statistical demand tool pretending to know what part of the market we cover and what part we do not cover depending on how much we produce, in this way we can make decisions regarding the quantity to produce considering the uncertainty put in the model through statistics. These results were validated using the Arena simulator for the current model of the company under study and the model proposed in the research work. The investment for the proposed model would be 16,467 USD, and the financial analysis shows that the project is profitable, since the IRR (54%) is higher than the COK (8.8%), in addition to having a positive NPV. It can be concluded that the proposed model has a positive impact on the results of the company under study.