Prosthetics presents a challenge for engineers, they must balance user needs with practicalities. In the past the compromises made have been towards simpler devices. Recently this has begun to change, new microprocessors, powerful motors and novel batteries make it possible to conceive of different solutions. These papers give an overview of what is now possible.

Winfried Heim C.P.(c), Bloorview MacMillan Children's Centre, Prosthetic and Myoelectric Services, 350 Rumsey Road, Toronto, Ontario (Canada) M4G 1R8.

We studied the pressure patterns in the residual limbs of transradial amputees during their voluntary commands for finger taps. Topographic maps of pressures exerted against the hard prosthetic socket were registered with an array of 32 pressure sensors, to produce residual kinetic images (RKIs) of the limb. Results with 2 untrained subjects demonstrated that RKIs are reliable decoders of efferent commands. Coupled with a trained filter, RKIs can provide biomimetic control over multiple degrees of freedom.

Rotational data in the form of measured three-dimensional rotations or orientations arise naturally in many fields of science, including biomechanics, orthopaedics and robotics. The cyclic topology of rotation spaces calls for special care and considerations when performing statistical analysis of rotational data. Relevant theory has been developed during the last three decades, and has become a standard tool in some areas. In relation to the study of human kinematics and motion however, these concepts have hardly been put to use. This paper gives an introduction to the intricacies of three-dimensional rotations, and provides a thorough geometric interpretation of several approaches to averaging rotational data. A set of novel, simple operators is presented. Simulations and a prosthetics-related real-world example involving wrist kinematics illuminate important aspects of the results. Finally generalizations and related subjects for further research are suggested.

In this paper a new prosthetic hand is presented that closely approximates the grasping abilities of a human hand. A large variety of different objects can be grasped reliably and the movements of the hand appear to natural. This five-finger hand has 15 degrees of freedom driven by small sized flexible fluidic actuators. The drives are within the fingers allowing a very compact and lightweight hand. Also, a concept for the control of different grasp types is presented. The characteristics of the new hand are illustrated.

The design of multifunctional upper limb prosthetics has been investigated in recent years. In 1998, the MANUS consortium was set up on the basis of multidisciplinary skills of the partners in order to provide a global approach to the problem of upper limb prosthetics. In the framework of this approach, the consortium addressed the development of appropriate hand mechanics, the analysis and implementation of control electronics, the development of advanced control strategies and the implementation of broad band command interfaces based on EMG.

As a result, the prototype of a multifunctional prosthesis was developed, integrated and clinically evaluated. This paper describes the objectives, technological approach and results of the MANUS prosthesis prototype.

The design of multifunctional upper limb prosthetics has been investigated in recent years. Several areas or research need to be developed for successful implementation of dextrous upper limb prosthesis, in particular better EMG interfaces for implementing command languages.

This article introduces a novel three-bit EMG command language concept as a user interface for the multifunctional MANUS prosthesis prototype. Following the global approach proposed by MANUS, a training process and the supporting Virtual Reality training software are also presented. The last section describes the evaluation process and the results.

This study presents a novel means of assessing upper limb tasks by using mechanical energy. Potential energy quantifies six work related activities, studied for 20 working-age non-prosthesis users and three powered below elbow prosthesis users. Two marker trajectories on each of the upper arms, forearms, and hands were captured using a 3-camera VICON 140™ system. Task and arm dominance of non-prosthesis users are highly significant ($p\,{<}\,0.01$) with arm dominance effects being more pronounced for prosthesis users. Qualitative inter-repetition consistency is also concordant with observed increased cumulative trauma disorders among prosthesis users.

Inverse dynamic calculations are often performed in the analysis of human ambulation. However, assumptions are made in the construction of the models that may not be valid for some prosthetic componentry, specifically polycentric knees used by transfemoral amputees. Data were collected for one subject using a polycentric knee. Moments and forces were calculated when ignoring the extra linkages of the knee and when including the extra linkages of the knee. Significant differences were seen in the hip moment and power between the two cases and indicate that care must be taken when comparing componentry that violates the basic assumptions of the rigid link model.

The aim of the project has been to provide a prosthesis which is capable of adapting to different modes of locomotion whilst optimising the hip power available to the user.

The prosthesis provides stance control ranging from minimal resistance to a yielding lock, capable of detecting level walking, ramp descent, stair descent, standing and instances of stumble. The stance resistance is set to preprogrammed levels for each mode, which matches the user's level of control.

10 amputees participated during the early development trails of the Adaptive Prosthesis for project validation. 7 used the limb for over 3 months and 3 for over 2 years. Early feedback indicated enhanced control and increased comfort and safety during ambulation and manoeuvring various terrains. Minor modification to increase reliability of cabling was also made.

In this paper the concept of a serial-parallel macro-milli manipulator is presented for medical applications by investigating the feasibility of combining two different robotic structures into one system. The combination of the two structures can be used as a redundant system in which stiffness, singularity and accuracy can be optimized as related to workspace demands and restrictions that are required for a priori selected medical applications. A case study has been experienced at LARM (Laboratory of Robotics and Mechatronics) at the University of Cassino through a new system composed of a commercial Adept COBRA and CaPaMan2bis (Cassino Parallel Manipulator 2bis). The performance of this system is compared with a previous serial-parallel macro-milli robotic system also developed at LARM in Cassino. Experimental tests have been carried out in order to characterize the proposed systems.

The complexity and uncertainty of the cross-sectional shape of the parts to be mated is one of the main reasons that misalignment between them occurs in assembly processes. Misalignment cannot only give rise to assembly failure but also cause damage to the parts or a robot due to large contact force. Therefore, misalignment sensing and compensation is essential for successful assembly operation. In this paper, we propose a novel misalignment estimation and compensation method which does not need any advance information on the cross-sectional shapes of the mating parts. This method utilizes a $\varphi-r$ transformation and an M-estimation pattern matching technique with misalignment images of a peg and a hole taken by an omni-directional visual sensing system during assembly. At every sampling instant during assembly action, it furnishes information on the relative position and orientation between the mating parts, and thus helps to estimate and compensate any possible misalignment between them. Also, a series of experiments are performed with a couple of peg-in-hole tasks, and the results are discussed. The experimental results show that the proposed method is effective for misalignment compensation in robotic assembly even though there is no prior information on part geometry and the images are very noisy.

In this paper an algorithm for real-time estimation of the position and orientation of a moving object using a video camera is presented. The algorithm is based on the extended Kalman filter which iteratively computes the object pose from the position measured in the image plane of a set of feature points of the object. A new technique is proposed for the selection of the optimal feature points based on the representation of the object geometry by means of a Binary Space Partitioning (BSP) tree. At each sample time, a visit algorithm of the tree allows pre-selecting all the feature points of the object that are visible from the camera in the pose predicted by the Kalman filter. A further selection is performed to find the optimal set of visible points to be used for image feature extraction. Experimental results are presented which confirm the feasibility and effectiveness of the proposed technique.

In this paper, an alternate passive leg structure is proposed for the Tricept machine tool to form a modified Tricept machine tool. The global stiffness of the modified Tricept machine tool is derived and compared with that of the Tricept machine tool. First, the configurations of the Tricept machine tool and the modified Tricept machine tool are introduced, respectively. Then, the global velocity equations are derived and the stiffness models of the two configurations are presented and analyzed. Finally, the advantages and disadvantages of the two types of passive leg structure are analyzed and discussed and stiffness simulations are conducted.

By using the impedance perception algorithm, which we previously proposed, the stiffness matrix that constrains the motion-force relation of a robot's end-effector is calculated in real-time. In this paper, we describe methods of extracting information concerning certain local properties of object surfaces from the stiffness matrix when the end-effector is slid on the surfaces. We develop two algorithms to deal with two classes of surfaces: flat and convex cylindrical surfaces. Since the proposed methods are designed separately from control strategies, they can be used for both autonomous and remote controlled robots. The results of preliminary experiments are presented.