Lot sizing and scheduling problem is one of the most important problems in production systems which has a direct influence on financial profit or loss and customer satisfaction of production companies. In many of previous studies, single-speed-assumption for machines was considered. Moreover, many researches in the literature have used exact deterministic data and have rarely considered environmental issues and the environmental impact on production planning problems. However, all mentioned items are important and influential in these problems. In this research, the impact of mentioned items on lot sizing and scheduling problem in flow shop is considered. The speed of a production machine influences production, backorder, and holding cost and also affects scrape rate and carbon monoxide emission. For meeting the demand of a period, we can increase the speed of machines during that period so as to let manufacturing cost, carbon monoxide emission, and scrape rate increase. In this way, we would decrease backordering cost and meet the customer’s satisfaction.In the real world, the input data, especially demand which is the most important one, is stochastic and multiple variables such as forecast error and economic issues are influential in forecasting the demand. Therefore, utilizing stochastic optimization methods to provide a feasible production plan close to the optimal solution is desirable. In this research, two mathematical models for the problem with above-mentioned assumptions and in the deterministic environment was presented. Solving 400 random examples revealed that the second model is more efficient compared to the first one; then, to face stochastic environment the Bertsimas-Sim and Alem-Morabito Robust optimization approach was used. For analyzing the influence of uncertainty and the impact of uncertainty range, 1800 random examples were solved showing that increasing length of the uncertainty range and rising the uncertainty coefficient, individually or together, will cause higher production costs. In this research, the presented robust optimization model is complicated and of NP-Hard ltr"
