Page 66 - Biomimetics : Biologically Inspired Technologies

Bar-Cohen : Biomimetics: Biologically Inspired Technologies DK3163_c002 Final Proof page 52 21.9.2005 11:39am

52 Biomimetics: Biologically Inspired Technologies

2001), smart materials that can remodel (Anderson et al., 2004) and be fabricated using molecular
self-assembly (Zhang, 2003). If these concepts can be integrated in a system that allows for
transport of the necessary components and removal of the waste products then remodeling polymer
actuators may become available in the future.
The use of sarcomeres as the basic functional unit also imposes limitations on the functionality
of muscles. It is likely that millions of years of evolution have resulted in a full exploration of the
sarcomere design. Consequently, it seems unlikely that the design has the potential to generate
2
tensions far above the 2200 kN/m , strain larger than 200% or shortening rates above 40 lengths
per second. These performance metrics are eventually limited by space requirements and the speed
of enzyme actions. For example, to accommodate the large forces generated by the claw closer
muscles of the crab, the thickness of the myosin filament has to increase. As a consequence, there is
less space for the actin filaments. This will limit the maximum amount of crossbridges in a certain
volume and thus the specific tension. Understanding these limitations may be useful for the design
of future actuators.
Mimicking the sarcomeric design of muscle in a synthetic muscle may prove to be a first step
towards a novel class of robust and functionally diverse actuators, and initial attempts look
promising (Frank and Schilling, 1998). The next step will require an integrative systems approach
to understand and mimic the functions of biological musculoskeletal systems during natural
movements (Full and Koditschek, 1999). This approach will identify the biomechanical principles
to be introduced in artificial models. An integrated approach to artificial muscle design has a strong
research potential. As an example, realistic biomechanical models of human limbs for analysis of
locomotion, with emphasis on understanding the underlying geometries and control problems,
provide an interesting basis to conceive a systems-based approach: large groups of muscle tendon
complexes have been successfully modeled as simple contractile elements in a functional model
(Roberts and Marsh 2003); redundancy problems associated with large muscle numbers are solved
with the proper control criteria (Rehbinder and Martin, 2001). Most importantly, the qualitative
insight obtained from models of biomechanical and control mechanisms are to be included in the
design of novel biomimetic muscular systems.
A biomimetic muscle must be provided with versatility and adaptability; with current state-of-
the-art actuator technology and its known limitations, this can be obtained if conceived in an
integrated approach. Examples of this can be found in novel applications in the field of biorobotics
and prosthetic devices. For example, a force-controllable ankle joint actuator for an ankle–foot
orthosis (Blaya and Herr, 2004) conceived as combination of controllable devices (DC motor,
springs) with an adaptive control strategy defined upon the biomechanical model of the anatomical
joint, can result in an actuator system that can adapt dynamically partially recovering a specific gait
disorder (drop foot) suffered by a group of patients. The joint impedance control introduced through
the series elastic actuator reduces significantly the foot slap and improves swing phase dynamics in
patients, as reported by the authors. Crucial constructive needs expected for such a system — and
any biomimetic wearable device — are low volume and size, low energy consumption, quiet
operation, low heat dissipation, and high torque (i.e., 3.3 W per body kilogram are required at the
beginning of the leg swing). These challenges are to be overcome by new actuators and materials,
providing lifelike characteristics. The weak musculoskeletal system in this case not only requires
assistance to control the impedance but also power generation (peak demand during gait, 3.3 W per
body kilogram) and other compensations to avoid other disorders found under the same muscular
disabilities, like dragging of the toe during swing phase, incomplete forefoot rocker and difficulty
to raise the foot. Such a biomimetic system can increase its level of functionality by increasing
the level of system integration. Following this example of biomimetic actuation and control for
orthopedics, a novel system for the impaired lower leg is being developed (Moreno et al., 2004). It
includes elements imitating the roles of anatomical parts, like tendons to assist powered acceler-
ations or the roles of biarticular muscles in a limb (Hof, 2001) to include the coordination
mechanisms.