Gesteme-free context-aware adaptation of robot behavior in human-robot cooperation.

BACKGROUND: Cooperative robotics is receiving greater acceptance because the typical advantages provided by manipulators are combined with an intuitive usage. In particular, hands-on robotics may benefit from the adaptation of the assistant behavior with respect to the activity currently performed by the user. A fast and reliable classification of human activities is required, as well as strategies to smoothly modify the control of the manipulator. In this scenario, gesteme-based motion classification is inadequate because it needs the observation of a wide signal percentage and the definition of a rich vocabulary. OBJECTIVE: In this work, a system able to recognize the user's current activity without a vocabulary of gestemes, and to accordingly adapt the manipulator's dynamic behavior is presented. METHODS AND MATERIAL: An underlying stochastic model fits variations in the user's guidance forces and the resulting trajectories of the manipulator's end-effector with a set of Gaussian distribution. The high-level switching between these distributions is captured with hidden Markov models. The dynamic of the KUKA light-weight robot, a torque-controlled manipulator, is modified with respect to the classified activity using sigmoidal-shaped functions. The presented system is validated over a pool of 12 näive users in a scenario that addresses surgical targeting tasks on soft tissue. The robot's assistance is adapted in order to obtain a stiff behavior during activities that require critical accuracy constraint, and higher compliance during wide movements. Both the ability to provide the correct classification at each moment (sample accuracy) and the capability of correctly identify the correct sequence of activity (sequence accuracy) were evaluated. RESULTS: The proposed classifier is fast and accurate in all the experiments conducted (80% sample accuracy after the observation of ∼450ms of signal). Moreover, the ability of recognize the correct sequence of activities, without unwanted transitions is guaranteed (sequence accuracy ∼90% when computed far away from user desired transitions). Finally, the proposed activity-based adaptation of the robot's dynamic does not lead to a not smooth behavior (high smoothness, i.e. normalized jerk score <0.01). CONCLUSION: The provided system is able to dynamic assist the operator during cooperation in the presented scenario.

Redundancy optimization strategy for hands-on robotic surgery.

During hands-on cooperative surgery, the use of a redundant robot allows to address encumbrance issues in the Operating Room (OR), which can occur due to the presence of large medical instrumentation, such as the surgical microscope. This work presents a new Null Space Optimization (NSO) strategy to constraint the position of the manipulator's elbow within predefined range of motions, according to the spatial requirements of the specific procedure, also taking into account the physical joint limits of the robotic assistant. The proposed strategy was applied to the 7 degrees of freedom (dof) lightweight robot LWR4+. The performance of the NSO was compared to two state-of-the-art null space optimization strategies, i.e. damped posture and fixed optimal posture, over a pool of three non-expert users in both strict (20deg) and negligible (100deg) angular encumbrance limitations. The NSO strategy was proved versatile in providing wide elbow mobility together with safe distance from relevant continuity null space boundaries, guaranteeing smooth guidance trajectories. Future works would be performed in order to evaluate the potential feasibility of NSO in a real surgical scenario.

Recognition of user's activity for adaptive cooperative assistance in robotic surgery.

During hands-on robotic surgery it is advisable to know how and when to provide the surgeon with different assistance levels with respect to the current performed activity. Gesteme-based on-line classification requires the definition of a complete set of primitives and the observation of large signal percentage. In this work an on-line, gesteme-free activity recognition method is addressed. The algorithm models the guidance forces and the resulting trajectory of the manipulator with 26 low-level components of a Gaussian Mixture Model (GMM). Temporal switching among the components is modeled with a Hidden Markov Model (HMM). Tests are performed in a simplified scenario over a pool of 5 non-surgeon users. Classification accuracy resulted higher than 89% after the observation of a 300 ms-long signal. Future work will address the use of the current detected activity to on-line trigger different strategies to control the manipulator and adapt the level of assistance.

Hip joint centre position estimation using a dual unscented Kalman filter for computer-assisted orthopaedic surgery.

In computer-assisted knee surgery, the accuracy of the localization of the femur centre of rotation relative to the hip-bone (hip joint centre) is affected by the unavoidable and untracked pelvic movements because only the femoral pose is acquired during passive pivoting manoeuvres. We present a dual unscented Kalman filter algorithm that allows the estimation of the hip joint centre also using as input the position of a pelvic reference point that can be acquired with a skin marker placed on the hip, without increasing the invasiveness of the surgical procedure. A comparative assessment of the algorithm was carried out using data provided by in vitro experiments mimicking in vivo surgical conditions. Soft tissue artefacts were simulated and superimposed onto the position of a pelvic landmark. Femoral pivoting made of a sequence of star-like quasi-planar movements followed by a circumduction was performed. The dual unscented Kalman filter method proved to be less sensitive to pelvic displacements, which were shown to be larger during the manoeuvres in which the femur was more adducted. Comparable accuracy between all the analysed methods resulted for hip joint centre displacements smaller than 1 mm (error: 2.2 ± [0.2; 0.3] mm, median ± [inter-quartile range 25%; inter-quartile range 75%]) and between 1 and 6 mm (error: 4.8 ± [0.5; 0.8] mm) during planar movements. When the hip joint centre displacement exceeded 6 mm, the dual unscented Kalman filter proved to be more accurate than the other methods by 30% during multi-planar movements (error: 5.2 ± [1.2; 1] mm).

Hip joint centre localisation with an unscented Kalman filter.

The accurate estimation of the hip joint centre (HJC) in gait analysis and in computer assisted orthopaedic procedures is a basic requirement. Functional methods, based on rigid body localisation, assessing the kinematics of the femur during circumduction movements (pivoting) have been used for estimating the HJC. Localising the femoral segment only, as it is usually done in total knee replacement procedure, can give rise to estimation errors, since the pelvis, during the passive pivoting manoeuvre, might undergo spatial displacements. This paper presents the design and test of an unscented Kalman filter that allows the estimation of the HJC by observing the pose of the femur and the 3D coordinates of a single marker attached to the pelvis. This new approach was validated using a hip joint mechanical simulator, mimicking both hard and soft tissues. The algorithm performances were compared with the literature standards and proved to have better performances in case of pelvis translation greater than 8 mm, thus satisfying the clinical requirements of the application.

Design and milling manufacture of polyurethane custom contoured cushions for wheelchair users.

BACKGROUND: The design of custom contoured cushions manufactured in flexible polyurethane foams is an option to improve positioning and comfort for people with disabilities that spend most of the day seated in the same position. These surfaces increase the contact area between the seat and the user. This fact contributes to minimise the local pressures that can generate problems like decubitus ulcers. The present research aims at establishing development routes for custom cushion production to wheelchair users. This study also contributes to the investigation of Computer Numerical Control (CNC) machining of flexible polyurethane foams. METHOD: The proposed route to obtain the customised seat began with acquiring the user's contour in adequate posture through plaster cast. To collect the surface geometry, the cast was three-dimensionally scanned and manipulated in CAD/CAM software. CNC milling parameters such as tools, spindle speeds and feed rates to machine flexible polyurethane foams were tested. These parameters were analysed regarding the surface quality. The best parameters were then tested in a customised seat. The possible dimensional changes generated during foam cutting were analysed through 3D scanning. Also, the customised seat pressure and temperature distribution was tested. RESULTS: The best parameters found for foams with a density of 50kg/cm(3) were high spindle speeds (24000 rpm) and feed rates between 2400-4000mm/min. Those parameters did not generate significant deformities in the machined cushions. The custom contoured cushion satisfactorily increased the contact area between wheelchair and user, as it distributed pressure and heat evenly. CONCLUSION: Through this study it was possible to define routes for the development and manufacturing of customised seats using direct CNC milling in flexible polyurethane foams. It also showed that custom contoured cushions efficiently distribute pressure and temperature, which is believed to minimise tissue lesions such as pressure ulcers.