Page 193 - Biomimetics : Biologically Inspired Technologies

Bar-Cohen : Biomimetics: Biologically Inspired Technologies DK3163_c006 Final Proof page 179 21.9.2005 2:56am

Robotic Mechanisms 179

Largely a reapplication of information processing theory to biology (arising from the hybrid
discipline of bioinformatics), this systems paradigm of life as an integrated web of interaction at
many scales, has largely moved to center of the biological sciences, complementing the reductionist
view of the organism (Dickinson et al., 2000). This chapter adopts this paradigm to frame a biorobot
as an artificial organism — a me ´lange of composites, subsystems, and integrated supersystems,
which are assembled to achieve a desired function, or multiple functions as the case may be.
To be definitively a robot though, an artificial organism must be capable of motility or mobility or
both. Otherwise, the phrase ‘‘artificial organism’’ can apply to biorobot’s close cousins: artificial life,
computer animated agents, and the like. Physical embodiment, the ability to move, the ability to
grasp and handle objects, to locomote: these distinguish a robot. And yet clearly, the overlapping of
technologies makes it difficult, and almost arbitrary, to define ‘‘robotics’’ at this stage. Indeed, many
technologies are relevant to robotics function and are useful without mobile physical bodies, with
examples including machine vision, automatic speech recognition (ASR), and natural language
processing (NLP). Alternately, the robot-like body may exist perfectly well without intelligence, as
is the case with animatronics — entertainment robots used in themeparks and films. Truthfully,
many technologies relevant to robotics have yet to be integrated into robots or into whole synthetic
organisms. Such integration represents one of the grand challenges in biomimetic robotics.
As such integration inches forward, robots will reap benefits from the numerous bio-inspired
technologies that are flowering presently. Bio-inspired tunable optics (Buckley et al., 2004), bio-
inspired sensors, actuators (Bar-Cohen, 2002), and materials promise to boost the performance of
robots. Make no mistake, robots also benefit from technology advances that are not bio-inspired,
such as advanced manufacturing technologies like MicroElectricalMechanical systems (MEMS)
and shape deposition manufacturing (SDM).
Many of these technologies can be well used to emulate biological systems. For example,
advanced manufacturing technologies result in faster, iterative robot design (Bailey et al., 2000),
more complex yet affordable robots, and designs at meso, micro, and even nanometer scales.
Advanced digital visualization tools are enabling better design and finite element testing before the
expense and trouble of manufacturing. Numerous shifts in design paradigms can be associated with
the new tools of manufacturing, design, and the biological sciences.
Biological systems are more capable than technology in countless ways. And yet, technology is
capable in some ways that biology is not. Moreover, when performing the same task as an
organism, humans can often use very different strategies, for example, aerofoil wings versus
flapping wings to achieve flying. So, the most effective robotic creatures can be an amalgam of
pure technology and translated or emulated biology.
When emulating bio-systems, the design approach may be either directly imitative (biomi-
metics) or inspired by physical principles derived from the study of organisms (bio-inspired). In the
first approach, technology approximates the end result or function of an organ or organism, for
example: when rotary motors turn leg joints, the mechanics are very different from those of
biological leg muscles, though the effected end motion of the leg is similar. In the second approach,
the principles extracted from bio-systems may be applied in ways very much unlike those exhibited
in the originating organism; for example, the crystalline structures of starfish eyes are being used to
create far more efficient fiber-optic routing hubs (Aizenberg et al., 2001). The two sets are not
mutually exclusive — there are many derivations and combinations of these techniques, and the end
objective is similar to create machines that move and function usefully by transferring bio-systems
into techno-systems.
Truly, as physical nature manifested through biological evolution, it solved a staggering number
of design and engineering problems. The job of the bioroboticist, then, is to discover and to abstract
these solutions, and to implement them in the limited media of human technology.
Computers grow more powerful and cheaper. Materials are growing more diverse and effective
in many ways. Manufacturing technologies are growing more automated, efficient, effective, and
flexible.