I. INTRODUCTION
The development of new technologies has steered the
design of complex and advanced robotic systems, such
as exoskeletons, to achieve a high level of performance
at reasonable production costs. Therefore, the design and
construction of exoskeletons for rehabilitation, functional
assistance or to improve the quality of life has become
feasible [1]. The research and development of such kind of
robots binds technology and science with a common goal: to
produce a robust, safe, comfortable and economical assistive
devices inspired by nature. Nevertheless, unlike many other
engineering products, there is no systematic and commonly
established procedure to design exoskeletons.
*This work was supported by CNPq and FAPESP (2010/17181-0)
1Rafael Sanchez Souza is with Department of Mechatronics and Mechanical
Systems Engineering, Polytechnic School of University of Sao Paulo,
Sao Paulo, Brazil. rafaelsouza@usp.br
2Filippo Sanfilippo is with the Department of Engineering Cybernetics,
Norwegian University of Science and Technology, 7491 Trondheim, Norway.
filippo.sanfilippo@ntnu.no
3Jos´e Reinaldo Silva is with Department of Mechatronics and Mechanical
Systems Engineering, Polytechnic School of University of Sao Paulo, Sao
Paulo, Brazil. reinaldo@usp.br
4Arturo Forner Cordero is with Department of Mechatronics and Mechanical
Systems Engineering, Polytechnic School of University of Sao Paulo,
Sao Paulo, Brazil. aforner@usp.br
There are major challenges involved in exoskeleton design.
One of the most demanding tasks is that the system must
be able to reproduce human movements to a functional
level. This means more than seven degrees of freedom to be
controlled in upper or lower limbs [2]. In addition, it must be
able to intervene in the exactly desired way with the subjects
movement - to control the interaction force between user and
robot. Many groups have been researching these technologies
and some solutions have been proposed and implemented,
such as the exoskeleton ARMin II [3] and the Impedance
Control [4].
Yet, systems have functional - as the aforementioned - and
non-functional requirements. Although many works discuss,
amongst others, wearability, usability and cost [5], nonfunctional
requirements are still often disregarded during
technology development, as also is the case with exoskeleton
research. Therefore, the user acceptance represents a key
factor decision-making along the project development.
When designing such systems, modularity may be used as
a possible approach. For instance, an upper limb exoskeleton
prototype was developed in [6]. The mechanical design has
a particular feature; the segments can be divided into very
similar modules formed by a joint and a link. This modular
approach, while promising, has not been sufficiently explored
in exoskeleton research [7]. Modularity brings development
and maintenance costs down due to standardization of components,
increasing portability, adaptability [8] and robustness
[9]. Rapid prototyping may also be combined to obtain
faster results [10], [11]. With modularity, non-functional
requirements are also expected to be contemplated.