How do we structure manipulation planning so that tactile feedback is not only relevant, but also convenient? We propose to do it by targeting contacts with the object geometric-rich features (for estimation) and defining dynamic systems with simple mechanics and efficient planners (for control). We focus on the application of manipulating an object on a table-top from an arbitrary initial pose to an arbitrary target pose while being robust against 1) external perturbations and 2) uncertainty in object pose.
Contact: Francois Hogan (fhogan [at] mit.edu)Paper
Tactile Dexterity: Manipulation Primitives with Tactile Feedback
Francois R. Hogan, Jose Ballester, Siyuan Dong, and Alberto Rodriguez.
Video
Tactile Control
We divide the role of tactile control into two goals: 1) control the contact state between the end-effector and the object (contact/no-contact, stick/slip, forces) and 2) control the object state by tracking the object with a tactile-based state estimator.
Object State Controller
We track the object pose from the geometric features detected in the tactile sensors.
Contact State Controller
We monitor slippage between the object and the end effector to regulate the applied forces on the object and enforce a desired contact state.
Robust Primitives
The tactile controllers enable the development of robust primitives, with the ability to handle external external perturbations and deal with object pose uncertainty.
Pull Regulation
We track the object pose from the geometric features detected in the tactile sensors.
Grasp Regulation
We monitor slippage between the object and the end effector to regulate the applied forces on the object and enforce a desired contact state.
Primitives
Key to this formulation is the decomposition of manipulation plans into sequences of manipulation primitives with simple mechanics and efficient planners. This structuring of the manipulation task into simpler behaviors gives us the freedom to design interactions for which we understand the mechanics, are able to interpret the tactile information, and can develop effective planning algorithms.
Pivot
Grasp
Pull
Push
Manipulation Planning
We formulate the manipulation problem as a sequencing of primitive behaviors. Given a desired object transformations, it is unlikely that a single manipulation primitive will have the ability to achieve it. This is due to the fact that many primitives can only achieve limited object reconfigurations. To achieve general object transformations, manipulation primitives must be combined. We formulate sequencing of manipulation primitives as a graph search problem.