The increasing deployment of robots in factories demands cost-efficient and non-expert programming approaches. This project aims to develop an easy-to-use solution for controlling and creating robot tasks. The proposed approach utilizes intuitive and visual programming with Behavior Trees (BTs) and Reinforcement Learning (RL) for control, based on the principles of Skill-Based Systems (SBS). The focus is on applying this system in Material Acceleration Platforms (MAPs), specifically for Self-Driving Labs (SDLs). A Matrix Production System (MPS) with shuttles and manipulators served as a use case for validating the solution. Research showed challenges in automating lab procedures, such as task transfer and lab layout limitations, hinder the automation of lab work. The solution presented demonstrated capabilities of creating and executing BTs on the MPS and the RL agent successfully navigated obstacle-free environments but faced difficulties with multiple obstacles due to control and behavior tendencies. The system serves as a proof of concept but requires further improvements before being put to use. Overall, the novel combination of BT and RL for multi-robot systems in MAPs shows promise, in advancing the automation of lab tasks and material discovery.