Automated temporal planning is a critical area of Artificial Intelligence with applications ranging from robotics to flexible manufacturing. In the context of the STEP-RL project, funded by the European Research Council, the PhD candidate will research novel approaches based on reinforcement learning to guide automated temporal planners, specializing them for specific applications. Examples of learned knowledge that will be considered include pruning rules and macro-actions. Different kinds of plan quality metrics and diverse dynamics for numeric fluents will be considered and studied in this research.