A modular approach for remote operation of humanoid robots in search and rescue scenarios

In the present work we have designed and implemented a modular, robust and user-friendly Pilot Interface meant to control humanoid robots in rescue scenarios during dangerous missions. We follow the common approach where the robot is semi-autonomous and it is remotely controlled by a human operator. In our implementation, YARP is used both as a communication channel for low-level hardware components and as an interconnecting framework between control modules. The interface features the capability to receive the status of these modules continuously and request actions when required. In addition, ROS is used to retrieve data from different types of sensors and to display relevant information of the robot status such as joint positions, velocities and torques, force/torque measurements and inertial data. Furthermore the operator is immersed into a 3D reconstruction of the environment and is enabled to manipulate 3D virtual objects. The Pilot Interface allows the operator to control the robot at three different levels. The high-level control deals with human-like actions which involve the whole robot’s actuation and perception. For instance, we successfully teleoperated IIT’s COmpliant huMANoid (COMAN) platform to execute complex navigation tasks through the composition of elementary walking commands (e.g.[walk_forward, 1m]). The mid-level control generates tasks in cartesian space, based on the position and orientation of objects of interest (i.e. valve, door handle) w.r.t. a reference frame on the robot. The low level control operates in joint space and is meant as a last resort tool to perform fine adjustments (e.g. release a trapped limb). Finally, our Pilot Interface is adaptable to different tasks, strategies and pilot’s needs, thanks to a modular architecture of the system which enables to add/remove single front-end components (e.g. GUI widgets) as well as back-end control modules on the fly.