Page 211 - Biomimetics : Biologically Inspired Technologies

Bar-Cohen : Biomimetics: Biologically Inspired Technologies DK3163_c006 Final Proof page 197 21.9.2005 2:56am

Robotic Mechanisms 197

We seek to address remaining challenges that were identified with the use of SPEM. These
challenges include the need to find ways to remove the sacrificial matrix more effectively and
quickly. There is a need to optimize the pore structure for expressive robot faces. Also, it is
necessary to design the SPEM to interoperate with other mechanical systems, actuators and anchor
placement, and mechanical attachments such that the face moves truthfully. Once we have devel-
oped the required material and procedures, we will have to evaluate the robustness, stability, and
mechanical performance of SPEM materials.
Even more dramatically, the material requires less than 1/30th the force to compress relative to
nonporous casts of the same silicone material. In facial expression robotics, this is particularly
advantageous as all expressions both compress facial soft tissues and elongate facial soft tissues.
Figure 6.17 shows HER’s latest robot EVA, which has been rendered using several of the
silicone SPEM techniques described above. Because of the silicone SPEM, EVA requires a tiny
fraction of the force to move into facial expressions, relative to other animatronic materials. For this
reason, this robot’s 36 DOF will run for hours on four AA batteries, consuming less than 10 W
average and 40 W peak.
As of the time of this chapter’s writing, the author’s robots can be seen in action at the follow-
ing urls: http://ndeaa.jpl.nasa.gov/nasa-nde/biomimetics/Biomimetic-robot-Hanson.mov and http://
androidworld.com/HansonHead.wmv.
For widespread applicability, bio-inspired robotic materials, structures, and systems need to be
manufacturable inexpensively and in bulk quantities. The toolset for creating bio-inspired robots is
clearly maturing, though much work remains.

6.5 CONCLUSORY REMARKS

Clearly the future glows for bio-inspired robotics. Many trends are showing high degrees of
functionality, and yet are increasing rapidly in function: computational hardware, materials,
software, and mobility are examples. Yet, daunting quantities of work remain to create robots
that are as capable as animals or humans. In fact, the increasing functionality of biomimetic robots
and AI results in humbling insights regarding the complexity of the bio-systems, which let us know
how much we yet know about life.
Future work includes the improved software integration that would accelerate the functionality
of social robots and the advancement of automated design and prototyping systems for robotic
systems from macro scale facial expressions and locomotion systems, to micro scale actuation and
electronics. Additionally, the biosciences need to further discover what makes animals so effective,
and engineers need to replicate these discoveries in technology. As the economy of biorobotics
continues to expand, largely bolstered by the ongoing trends of increasing functionality, there
should be ample resources for future research in this exciting field.

Figure 6.16 Rectangular-celled SPEM made via rapid prototyping. Elongated state shown on the right. The
sample is 2 cm in width, 10 cm length, and 0.6 cm depth.