A novel linear-integral TOPSIS approach to mars mission simulator planning at NASA

Decision-makers are often subject to cognitive biases when making complex decisions. One such bias is apophenia (or pareidolia concerning visual information), where people perceive patterns that do not exist. Everyone experiences this phenomenon to varying degrees. Decision-makers often rely on cognitive shortcuts (heuristics) that deviate from Euclidean distances, perceiving relative rather than absolute distances in estimating the proximity of objects in multi-attribute decision-making. For instance, if a trajectory closely resembles a reference trajectory, the brain may perceive it as closer, even if another trajectory is technically closer in Euclidean terms but follows a more divergent path. Euclidean distances are used to assess the proximity between alternatives in the conventional Technique for Order of Preference by Similarity to the Ideal Solution (TOPSIS). However, this approach may not align with how the human brain processes complex information due to the limitations of Euclidean measures in capturing cognitive biases. Therefore, we propose a new linear-integral TOPSIS approach that actively integrates this human tendency to perceive patterns, enhancing its compatibility with cognitive biases. As the decision maker processes information, the method becomes influenced by the sequence dependencies among the attributes. We demonstrate that this issue can be linked to a specific version of the traveling salesperson problem when determining the best and worst sequence orders for processing an alternative’s attributes. To address this, we rely on the sequence induced by the preference structure and solve an open traveling salesperson problem to substantiate the robustness of the resulting sequence. We demonstrate the applicability of the new linear-integral TOPSIS to the Mars mission simulator planning at NASA’s mission directorate at Johnson Space Center.