Journal club: Bowyer et al 2013.
Bowyer et at. presents a review of active constrains in a generalized active constraint implementation framework; by defining, evaluating and enforcing these constraints. Geometry based definitions can simple as single point, and more complex e.g. a parametric 3D surface. The evaluation is made by computing metrics to and from the active constraint, to the constrained tool geometry (CTG) that can be on the tip of the robot, a section of joints or the complete device. To enforce the constraints several approaches are reviewed that range from a simple function of constrained proximity to a constrained joint optimization approach that considers individual links and end effector.
Discussion
Two-arm collaborative robots in minimally invasive surgery need fast, secure and seamless performance to reduce the workload and decrease operation time; active constrains aims to aid the user in reducing cognitive load, while considering hardware limitations. This by reducing the number of degrees of freedom to control, in the case of the envisioned setup: two collaborate arms attached to a main body, a setup that resembles endoscopes. Aspects on the use of active constraints within the two-arm structure are discussed: i) Envisioned constrains are equal for both of the arms? and why? ii) What kind of representation, evaluation and enforcement is the best given the comparison proposed? iii) Dynamic constraints will be less intrusive to the user? Why? iv) Evaluation metrics for the user cognitive load can overwrite defined constrains?
And subsequent future work is address by: i) Describing constrains in terms of the tools: static and dynamic for the body with enforcement and the instruments with limitation ii) Lines and parametric surfaces for simple testing (e.g. in case of depth preservation w.r.t. a frame) by proximity tracking and enforcement of the constraints iii) Tests needs to be carried out to ensure dynamic constraints are suitable from a users opinion iv) Registering an experiment with potential user for baseline, constrained body, constrained body and tools; and compare results.