Login
- Title:
The environmental impact of switching from deterministic to stochastic modeling in Sales & Operations Planning under uncertainty
- Document type:
- Nicht veröffentlichter Vortrag
- Author(s):
- Greil, T.; Grunow, M.
- Abstract:
- Companies today still employ deterministic methodologies in most of their supply chain planning processes. However, the same companies increasingly express their interest in advanced planning functionality to better handle uncertainty (e.g. probabilistic planning). We reason that adopting such more complex planning approaches might create differing environmental effects – either benefit or harm. Therefore, we investigate a sequential decision-making setting for Sales & Operations Planning in which a profit-maximizing firm switches from a deterministic to a stochastic approach. We model the deterministic status quo as a linear program with exogenously defined safety stock – similar to how this part of the Sales & Operations Planning process is carried out in practice. The stochastic approach is modeled as a Markov Decision Process to capitalize on incorporating the value of replanning under uncertainty. For larger instance sizes, we implement a Reinforcement Leaning approximation to obtain a solution. We then compare the policies in a simulator setting – both regarding achieved profit and the associated profit-oriented environmental impact. The environmental impact is merely determined (i.e. not optimized or constrained) to illustrate the worst possible situation in practice at companies today. We argue that the relative difference in environmental impact can greatly deviate from relative change in profit between the two approaches. Thus, switching to stochastic policies might increase profit only slightly but lead to a disproportionally lower (or higher) environmental impact. We furthermore investigate how the revealed forecast data (i.e. the approaches’ input) and the variability of the underlying stochastic process itself can affect the environmental impact.
- Event:
- 4th Conference of the EURO Working Group on Sustainable Supply Chains
- Publisher address:
- Hagen
- Year:
- 2023
BibTeX