Humanoid robots are designed to mimic humans’ physical appearance and intricate movements. The quest for creating humanoid robots with human-like motion capabilities represents a fundamental milestone in robotics. On one side, researchers focus on providing robots with the same ability as humans to perform multiple tasks simultaneously. On the other hand, researchers focus on designing dynamic controllers that enable humanoid robots to run or jump at different elevations, improving their mobility in unstructured environments.

In this context, the HiFliTE project aims at making two-legged robots able to perform tasks not feasible for the robot structure (e.g., because of height, joint position limits, lack of limbs) by exploiting the flight phase of a jump. To achieve this goal, a high-level control framework will be designed to plan the motion during the phases of take-off, flight, and landing. During the flight phase, the goal will be controlling one end-effector of the robot to perform a defined task while the rest of the body compensates for the end-effector motion and stabilizes the robot. This framework will be validated on a bipedal robot that is not equipped with arms to perform the task of opening the door. Since the robot has no arms, during the flight phase, one leg will be controlled to kick on the door handle/panic bar, and the other one and the torso to compensate for the motion.






MSc PhD Grazia Zambella


2024 - 2025