Master Thesis

Explaining robotic planning decision points along execution

Work default illustration



  • If you are interested in the proposal, please contact with the supervisors.


Explanation of robotic behaviours has been proved to be very important to improve the understanding of the users of such robots, which makes their trust in the robotic system improve.

This project explores the extension and reuse of the SoPS algorithm [1] to enable a robot to explain the different decision points or branches in a plan that is being executed. Here, the idea is that the planning domain will have different elements that change the cost of the action, without constraining the applicability of the actions. An example of this, which is also used in the SoPS paper [1], is that of preferences. In this case, the planner decides which action to take based on costs that may not be obvious to the user. Thus, by analyzing the space of plans (or a subset of it), the method may be able to describe why an action was taken by the planner, as the SoPS tree allows the comparison of the different plans at branching points based on the maximum achievable reward of each plan.

[1] G. Canal, C. Torras, and G. Alenyà, “Generating predicate suggestions based on the space of plans: an example of planning with preferences,” User Modeling and User-Adapted Interaction, vol. 33, pp. 333–357, 2023.

The work is under the scope of the following projects:

  • COHERENT: Collaborative hierarchical robotic explanations (web)