Integrated visual servoing and force control. The task frame approach. (Q1421037)
From MaRDI portal
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Integrated visual servoing and force control. The task frame approach. |
scientific article |
Statements
Integrated visual servoing and force control. The task frame approach. (English)
0 references
25 January 2004
0 references
The goal of this work on robotics (which arose from the dissertation of the first author) is to present ``a real-time approach\dots[for] integrating visual servoing and force tracking based on the task frame formalism in a hybrid control structure'' (page one). The notion of ``task frame'' central to this book is not well defined mathematically: ``Each direction \((x,y,z)\) is considered once as an axial direction and once as a polar direction.'' (?) ``Force, position and vision sensors can each command separate task frame directions'' (page 17). The authors continue using the expression ``task frame', which is still undefined. The notion is however qualified before (page 12); it is an ``orthogonal frame attached to the tool'' (the robot). Also one reads on page 18 that five kinds of directions can be considered: velocity, force, vision, tracking and feedforward. The reviewer did not manage to understand their associated explanations. Concretely, the authors want to use force and visual information to maintain the contact (with a force configuration) between a robotic manipulator and an object. They develop a simple method for distinguishing the object from a pixel image using a threshold and an associated measured area (this presumably needs more assumptions than the authors are giving); from this measure (cf. the mathematical meaning of this word) they try to associate contours and shapes. This with forces is used as a control input for regulation. A chapter details contour following of ``continuous curves'' (the authors certainly mean differentiable curves). The next one is a pendant when kinks are present. The authors use their jargon although it would not have been a lot of effort to explain the words they are using: ``dynamic look-and-move'', ``eye-in-hand'' (such a funny expression!). The index is not helpful. Many times the reviewer had trouble understanding what the authors write. The idea of this work is nice and stimulating, but there is certainly still a lot to do ahead.
0 references
robotics
0 references
visual servoing
0 references
force tracking
0 references
task frame
0 references
hybrid control
0 references
contact
0 references
robotic manipulator
0 references