Surgical tool pose estimation has been proven to be useful for high- and low- level feedback tasks including safety-enhancement, semantic feedback and surgical skill assessment. Tool pose estimation using monocular camera input is a well-studied research problem as the monocular camera is one of the ubiquitous sensor across the spectrum of robotic devices. Current state-of-the art methods for visual tool pose estimation are computationally expensive and require elaborate geometric and appearance models of surgical tools. We propose a visual tool pose estimation method that maps the visual bounding box to the 3D tool pose without any explicit knowledge of tool geometry using Gaussian process regression. The proposed approach can be generalized to any surgical tool and provides tool pose estimates with a variance estimate in real-time. We demonstrate rigorous evaluation of the method under various conditions that might effect the estimation process. In order to evaluate the algorithm, we have instrumented a standard box trainer kit with two laparoscopic tools to get simultaneous ground truth pose and a video feed.

In this paper, we generalize our random matrix based (RM-based) uncertainty model for manipulator Jacobian matrix to the dynamic model of the robotic systems. Conventional random variable based (RV-based) schemes require a detailed knowledge of the system parameters variation and may be not able to fully characterize the uncertainties of the complex dynamic systems. However, the proposed RM-based approach provides a probabilistic framework for systematic characterization of the uncertainties in the complex systems with limited available information. Moreover, RM-based uncertainty model is an efficient mathematical tool that ensures the kinematic and dynamic consistency and takes into account the system complexity, configuration, structural inter-dependencies, etc. The application of the RM-based uncertainty model is investigated using an example of kinematically redundant planar parallel manipulator (3-(P)RRR). The simulation results are compared with those obtained through conventional RV-based approach and the effectiveness of the proposed method is discussed.

High mobility, maneuverability and obstacle surmounting capabilities are highly desirable features for rough-terrain locomotion systems. In past work, we examined kinetostatic optimization of candidate articulated leg-wheel subsystem designs (based on the four-bar mechanism) for enhancing locomotion capabilities of land-based vehicles. Our goal was to: (i) achieve the greatest motion-ranges between wheel axle and chassis while (ii) reducing the overall actuation requirements by spring assist. In the current work, we examine the possibility of enhancing this terrain-accommodation by: (i) proposing an “adjustable four-bar” articulated-leg-wheel subsystem; (ii) with active-structural control to actively change subsystem parameters during the terrain traversal. Multiple leg-wheel design- parameters can affect the peak-static torque requirements as well as dynamic-bandwidth requirements for the leg-wheel actuation. The presented results compare and contrast online-structural reconfiguration planning (with multiple alternate active-adjustments) to reduce actuation requirements for a predetermined/sensed terrain traversal profile.