Five active axes artificial arm: a dedicated tool to collect body-segments relative movements and develop an effective algorithm for movement control

Designing a system to drive and control the motion of an artificial upper limb becomes more complex as the number of active axes of the arm increases. At present total upper limb prostheses have generally not more than three active joints (elbow, wrist and hand) which are directly driven by a couple of myoelectric signals. In general the joints are activated, one at a time with a sequence chosen by means of a mechanical or myoelectric switch. This approach produces quite unnatural movements and cannot be used when the number of the driven axes increases, as in our case where the presence of the shoulder adds two active axes. Furthermore the human brain is not able to control simultaneously all muscles suitable to produce the driving signals for all the axes. Thus, our task is to move our new five–d.o.f. artificial arm asking to the patient only to try to locate the chain end-effector (the hand) paying no attention to the joints coordinates, that will be automatically computed by the system. Therefore two main problems must be solved: the first one is to choose a way to tell the system where the patient will put his/her artificial hand and the second one is to synthesize the laws of motion for all active axes in order to obtain fairly natural movements. In this pilot work we employ myoelectric signals only for the hand open-close task, while we drive the hand displacements by means of little relative movements between other body segments. The current part of the research concerns the first problem and specifically the development of a tool for the synthesis of the algorithm to drive the hand. The tool is made of hardware to measure and collect the relative kinematics (e.g. of head and shoulder versus trunk) and software to build and test the driving algorithms based on recorded parameters processing. The apparatus must be portable and simple in order to be used for the preliminary tests not only in laboratory, but also in hospital or at patient home. Another aim of this work is to identify the minimum number of kinematic parameters that must be collected and the level of precision required, in order to choose the miniaturized measuring system (instrumented linkage, accelerometric system, set of biaxial commercial goniometers, etc.) that will equip the upper limb prosthetic system for the next clinical tests.