Operation scheduling on MDRS under uncertainty: comparing classical and robust models

We will consider the problem of scheduling a set of (experimental and logistic) operations in a constrained context such as the MDRS. When taking uncertainty into account while designing models, a robust solution that minimizes the probability of a failure in the mission global objectives can be obtained.

Based on real-life operations at MDRS, we will investigate different mathematical formulations for both deterministic and robust stochastic modelings of the problem, the latter taking uncertainty into account. In order to solve the deterministic problem, optimal and heuristic algorithms will be designed. Our principal goal is to show how the latter can be modified in order to take stochasticity into account while maximizing the probability of mission completion. Finally, computational experimentations will compare both approaches and highlight good practices for the benefit of future operation planning on Mars and in similar environments.