D'Claw platform supports three benchmarks --
Screw -- that are inspired by the fundamental behaviors observed in manipulation.
D'Claw Pose: Conform to the shape of the environment
Pose task is motivated by the primary objective of a manipulator to conform to its surrounding in order to prepare for the upcoming maneuvers -- commonly observed as various pre-grasp and latching maneuvers. This set of tasks is posed as trying to match randomly selected joint angle targets. Successful completion of this task demonstrates the capability of a manipulator to have controlled access to all its joints. Such control is a prerequisite for more advanced contact rich maneuvers to follow. This set of tasks are comparatively easier to train, thereby facilitates fast iteration cycles and a gradual transition to the rest of the tasks.
D'Claw Turn: Rotate to target angle
Turn task encapsulates the ability of a manipulator to reposition unactuated DoFs present in the environments to target configurations -- commonly observe as turning various knobs, latches and handles. This set of tasks is posed as trying to match randomly selected joint angle targets for the unactuated DoFs. Successful completion of this task demonstrates the ability of a manipulator to bring desired changes on external targets. Not only co-ordination between the internal DoFs but also understanding of environment dynamics, perceived via contact interactions, is required to succeed.
D'Claw Screw: Continuously rotate the unactuated object
Screw task focuses on the ability of a manipulator to continuously rotate an unactuated object at a constant velocity. This set of tasks is posed as trying to match joint angle targets that are themselves moving. Although very similar to turn tasks but the nuances of moving target challenge the manipulator's strategy to constantly evolve as the target drifts. Fingers often enter singular positions as the rotation progresses. A successful strategy needs to learn finger co-ordinated gating to simultaneous progress as well as stay out of local minima.