In search of a complete understanding of a joint’s function, one must understand both the anatomic parameters and how the brain controls the joint’s actuation. Accurate measurements of anatomical parameters are critical to non-linear biomechanical modeling and control and also to a clinical understanding of orthopaedic reconstruction. Likewise, new frontiers in the study of neuromuscular control contribute to our understanding of joint structure and function. One approach to study joint function is to use a joint simulator to actuate cadaver limbs. Towards the goals of understanding and improving human elbow joint control, a physiologic elbow joint simulator was previously constructed in our laboratory. It is the first elbow simulator to operate completely under closed-loop control. The closed-loop force control used to study joint mechanics permits measurement of moment arms in cadaveric elbow specimens. We hypothesized that the approach yields comparable results to previously-reported moment arm values.[1]

A fundamental principle of human motor behavior states that the accuracy of targeted movements relates reciprocally to their speed. This is quantified by Fitts’ Law, wherein movement time (MT) and index of difficulty (ID), the log 2 ratio of target distance (A) to target height (H) has logarithmic linear relationship; MT = a+b·log 2 (2A/H) = a+b·ID. The slope, b (seconds/bits), measures targeting performance as the time spent at each difficulty level, expressed as bits of information to be processed by the neuromotor system [1, 2]. Fitts’ paradigm is a common measure of the kinematic performance of the upper limb, but has not been applied to its dynamic performance. Herein, we developed a dynamic speed-accuracy trade-off (DSAT) test of grip force modulation, which can be used both for assessment and training.

A standard biomechanical testing protocol for evaluation of the sub-axial cervical spine is the application of pure bending moments to the free end of the spine (with opposing end fixed) and measurement of its motion response. The pure moment protocol is often used to compare spinal fusion instrumentation and has also been used to evaluate non-fusion instrumentation (e.g. disc arthroplasty devices) [1,2]. A variety of different testing systems have been employed to implement pure moment application. In cases where the loading is applied quasi-statically using a series of weights and pulleys the spine may relax between intermittent loading phases and/or unintended loading may be applied causing experimental artifact. Our objective was to use an existing programmable robotic testing platform (Spine Robot) to develop a novel real time force control strategy to simulate pure moment loading under precisely controlled continuous movement conditions. This would serve to advance robotic testing capabilities with an end goal to simulate different protocols in the same platform, and to potentially minimize fixturing and quasi-static artifacts.

The first study as regard with the application of robotic technology to the field of joint biomechaics was reported more than 20 years ago 1) . Since then, a variety of studies have employed commercially available articulated manipulators for the joint biomechanical studies 1–4) . However, such articulated manipulators are generally poor at stiffness and precision although they were basically designed to achieve high speeds of motion while performing tasks in a large work space. To solve the problem, we have previously developed a robotic system consisting of a custom-made 6-degree of freedom (6-DOF) manipulator and a universal force-moment sensor (UFS) 5) . Referring to the robotic system, the present study was aimed to develop a novel robotic system of rigid body/structure that allows a high-rate displacement/force control of the knee using a velocity-impedance control.