IntelliMan_WP5_Grasping, Manipulation and Arm-Hand Coordination_T5.4_Experience- and model-based Grasp Synthesis and Manipulation_WOSRA24

Automated planning is commonly used to obtain plans to solve particular tasks. To execute the plans, Behavior Trees (BTs) have emerged as a popular execution architecture due to their reactivity and modularity. Configuring the execution of a plan into a BT requires ex- panding high level actions into the proper BT structure. For this, the use of pre-defined rigid templates is common. In this work, in order to overcome the limitation of these methods, we present a new approach to describe robotic behaviors using ontologies and to adapt BTs to current specifications of the task and the world by reasoning on what modifiers to their base template need to be applied. These modifiers and their properties are formally defined using ontologies. We present a proof of concept of these modifiers applied to a base BT of a Pick operation.