CONTENTS

1.1 What Is Biologically Inspired Robotics?

1.2 History

1.3 Biologically Inspired Robot Design

1.4 Biologically Inspired Robot Control

1.5 Biologically Inspired Actuation and Sensing

1.6 Conclusion

References

1.1 What Is Biologically Inspired Robotics?

Biologically inspired robotics is an interdisciplinary subject of robotics and biology and consists of mainly two broad areas: biomimetics and bio-robotic modeling/analysis. Biomimetics draws inspiration from biology, and its primary concern is the application of biological ideas and phenomena to engineering problems in robotics. The topics cover almost every technical aspect of robotics including biologically inspired design, motion control, sensing, and actuation of robotic systems. A typical example of biomimetic robots is the humanoid robot, which is analogous with a human being in appearance and behavior (Figure 1.1). Bio-robotic modeling/analysis is the application of robotic models and principles to address biological issues such as recognition processes of the human brain, behaviors of animals and insects, etc For example, by using a model of a biomimetic robotic fish, it is possible to study the swimming dynamics of fish; it may be possible to model sensory motor control of human arms using a bionic arm.

1.2 History

Humans have tried to create mechanical systems that mimic the behaviors of animals and other living creatures for a long time The history can be traced back to development of the mechanical drink-serving waitress and musical players by Arab scholar and craftsman Al-Jazari in the thirteenth century and mechanical puppets or dolls such as the well-known Japanese karakuri ningyo in the eighteenth and nineteenth centuries. Probably the most famous example is the tremendous effort made to development of flying machines in the early twentieth century.

Designing robots that mimic animals and other living creatures dates back to the 1940s and 1950s (Beer 2009). The robotic tortoises developed by W. Gray Walter (Walter 1963) are most closely related to biologically inspired robotics. The tortoises are driven by motorized wheels and equipped with a light sensor and touch sensor. They are mobile robots indeed! When talking about the history of biomimetic robotics, it is necessary to note the work of Ichiro Kato's group at Waseda University in the early 1970s on design and control of biped robots (http://www.wikipedia.org/wiki/Humanoid_robot). They developed the first biped robot, WaBOT-1, in 1973 and a musician robot that played the piano in 1984. Their work laid the foundation for the research and development of present-day humanoid robots. Since the early 1980s, inspired by motion of snakes and spiders, Hirose and his group have designed several snake robots and legged robots (Hirose and Yamda 2009). Figure 1.2 shows the latest design of a snake robot created at the Shenyang Institute of Automation (China; Z. Liu et al. 2006). In 1997, Honda presented the first humanoid robot, Asimo, that truly has a humanoid appearance and integrates computer, control, sensing systems, power, and into a single stand-alone body (Hirai 1997) Since then, several humanoid robots, such as the Sony humanoid robot QRIO (Movellan et al. 2004), the AIST humanoid robot HRP-2 (Kaneko et al. 2004; Figure 1.1), and the BIT humanoid robot BHR-2 (Huang et al. 2005), have been developed in Japan, Europe, and China. The early biomimetic robots for entertainment include the robotic dog AIBO developed by Sony and the seal-mimetic robot PARO by Shibata (2004; Figure 1.3). With the advancement of sensing, actuation, and information technology, biologically inspired robotics is advancing rapidly with extensive study by robotics researchers and increasing investment from industries and governments worldwide. Different robots or robotic systems inspired by animals, insects, and fish have been developed. Figure 1.4 shows a robotic fish developed by Hu at the University of Essex (J. Liu, Dukes, and Hu 2005) Biomimetics has become one of the fastest growing topics in robotics in recent years.

1.3 Biologically Inspired Robot Design

Designing mechanisms for robots that mimic the motion of animals and other living creatures is one of the core problems in biologically inspired robotics The mechanisms of movement vary for different animals and other living creatures. Many mammals, such as cats, tigers, horses, etc., use four legs to move around, but humans rely on two legs to move Spiders use legs to climb, but snakes climb without legs The challenging issue here is how to realize biological movement using mechanical structures. Biological motion is generated by the interaction of muscles, joints, and tissues of a continuum deformable body There are no actuators that are as sophisticated as muscles, materials that are as soft as tissues, or joints that generate the complicated yet smooth motion that human and animal joints perform Therefore, it is crucial to develop simplified mechanisms that can generate motion similar to biological motion.

For example, a snake moves forward and backward using the frictional force between its body and the ground Its body is a deformable continuum whose geometric shape is used to control the frictional force The snake can change its speed by changing the shape of its body. Because it is difficult to design a mechanical structure that can deform freely and continuously by active control, existing robotic snakes employ a series of movable segments that are connected by a joint (Figure 1.5). Moreover, the snake relies on its skin to slide on the ground, and such skin cannot be made by current technologies, so wheels are attached to the segments If the connection joint allows the rotation about one axis, the robotic snake moves in a plane If the connection joint is a spherical joint that allows rotation about two perpendicular axes, the robotic snake can move in three-dimensional space; for example, to climb a tree.

By observing the motion of animals and insects, researchers have designed legged robots including biped or humanoid robots, four-legged robots that mimic the mechanisms of animal movement, and robots with eight, twelve, or even more legs. Legged robots move by repeatedly lifting and moving their legs backward or forward as animals do. Figure 1.6 shows a four-legged and a six-legged mobile robot. Other examples of biomimetic robots include the robotic fish developed at the University of Essex (see Figure 1.4).

1.4 Biologically Inspired Robot Control

The behaviors of animals and other living creatures inspire the development of new ideas for controlling the motion or behaviors of robots In principle, robotic control and biological control systems are similar They all work on the basis of sensory motor control. Biological systems are controlled by expansion and contraction of the muscles based on information collected by the biological sensors such as eyes, skin, ears, nose, etc. Robotic systems are controlled based on information feedback from robotic sensors using their actuators. The underlying principle for both robotic and biological systems is feedback control. Traditionally, researchers have designed control algorithms for robots using conventional methodologies and theories in controlling engineering. Different from traditional approaches, biologically inspired controllers are designed based on new philosophy inspired by biological systems Typical examples of biologically inspired approaches for robot control are behavior control, proposed by Brooks at the Massachusetts Institute of Technology (Brooks 1987); iterative learning control, developed by Arimoto et al. (1985); and intelligent control methods including genetic algorithms (Parker, Khoogar, and Goldberg 1989) and swarm control (Fukuda and Kawauchi 1990).

Behavior control was originally developed to solve the problem of motion control for mobile robots Traditionally, researchers have followed the procedure of sensing-perception-planning-control for controlling the motion of a mobile robot. In this method, the robot first uses its sensors to acquire information from the surrounding environment. Second, the acquired information is processed and interpreted As the third step, a motion is planned for the robot based on the information interpreted Finally, the robot executes the planned motion It was found that this sequential approach was not very useful for navigation of mobile robots because it took a lot of time to process and interpret the information and to plan the motion Behavior control employs the idea of reactive control, which is a typical behavior of biological systems. When physical stimulation is received by the human body, the body will immediately react to the stimulation. For example, when your hand gets too close to a hot stove, your hand will automatically move away from the source of heat once it feels hot. In behavior control, there are a number of primitive behaviors that are reactive responses to information collected by a sensor or a group of sensors in parallel (Figure 1.7). Primitive behaviors are arranged in a hierarchical structure so that coordination among them can be easily carried out. This approach is widely used in navigation control of mobile robots.

Iterative learning control was developed by Arimoto in 1982 based on observation of the human learning process of drawing lines, circles, and letters. When I was in primary school, whenever the teacher taught us a new Chinese character, he always asked us to write the character twenty or thirty times after class. The process of learning new characters is as follows: we first tried to write the character and then compared what we wrote (real trajectory in robotics) with the model (desired trajectory) so that we could correct our hand inputs the next time. By repeating this cycle several times, we could always write the character well Iterative learning control actually simulates this process Consider the case of a robot drawing a circle using iterative learning. Denote the desired trajectory of the robot in drawing the circle by (q_d (t), q_d (t), q_d (t)) and the real trajectory of the robot at the kth trial by (q¡t (t), q_k (t), q_k(t)). Let u_k (t) represent the joint input of the robot mani-plator at the kth trial. y_k (t) denotes the output of the robot at the kth trial, and y_d (t) is the desired output. The output could be either the velocity or the position of the robot. The input u_k+1 (t) of the robot at the next trial, that is, the (k + 1)-th trial, is given by the following P-type learning process:

u_k+1(t) = U_k (t) + K(y_d (t) - y_k (t)) (1.1)

where K is the positive-definite learning gain. The asymptotic convergence of the position error of the robot manipulator under iterative learning control has been proved using Lyapunov's theory. Figure 1.8 shows a block diagram of the iterative learning controller. Arimoto, Naniwa, and Suzuki (1990) also developed a D-type learning controller and a learning controller with a forgetting factor.

A genetic algorithm is a heuristic optimization method inspired by natural evolution of biological systems It uses evolution algorithms to search for the optimal solution. Genetic algorithms have been used in robotics for optimal control, planning, etc Swarm control is inspired by the group behaviors of ants A single ant has very limited ability to transport food. However, a group of ants can transport a large amount of food that is much heavier than the weight of a single ant. Swarm control is a decentralized and self-organized method for achieving collective behaviors of multiple robots, such as formation control, etc.

1.5 Biologically Inspired Actuation and Sensing

Sensors and actuators are major components of robotic systems. Development of robot actuators that function similar to muscles has long been a goal in the field of robotics. Muscles generate force by contraction of muscle fibers. The rubbertuator (Wang et al. 1992), developed by Bridgestone Corporation (Japan) in 1985, was the first commercial actuator that had similar characteristics to human muscles The rubbertuator is made from rubber tubes covered by braided fibers. By shortening and lengthening the rubber tubes as compressed air is fed in or blended out, it is possible to rotate a joint in a robotic arm The advantage of using rubber as the actuator is that it can control not only the position but the impedance/force of the robotic system. Unfortunately, the rubbertuator was not a successful product mainly because it was not suitable for many applications. In recent years, different artificial muscles, such as those using ionic polymer-metal composites (IPMCs; Kaneda et al. 2003; Oguro, Kawami, and Takenaka 1992), nano-tubes, and polyacrylonitrile, have been developed. The IMPC-based artificial muscle uses electricity to control its deformation The IMPC is made by coating a platinum or gold layer on an ion-exchange membrane, which is a perfluorosulfonic acid membrane. When an external voltage is applied to the metal layers, the ions of the ion-exchange membrane are attracted to the electrodes with water molecules. As a result, one side of the membrane will expand and the other side will shrink, so the IPMC bends at a high speed. The polyacrylonitrile artificial muscle uses the change of the pH value to contract. Artificial muscles have many potential applications such as in design of prosthetic limbs and in robotics, but tremendous efforts must still be made in design, modeling, analysis of characteristics, and control before the technology is mature.

Biologically inspired sensing involves two aspects: design of robot sensors based on biological sensors such as skin, eyes, ears, etc., and sensing biological signals of humans for robotic applications. Research on biologically inspired sensors includes efforts made in computer vision, in particular stereo vision, and development of tactile sensors, artificial ears and noses, etc. The work in sensing biological signals includes electromyography (EMG) and electroencephalography (EEG) measurements for robot control and human-robot interactions. Figure 1.9 shows the surface EMG signals measured when a hand grasps a ball. If it is possible to directly understand the intention of a person from his EEG/EMG signals, we can develop an EEG/EMG-based closed-loop control system for prosthetic limbs so that the artificial limbs could perform as well as the original limbs do. Similarly, by understanding the EEG signals generated by the brain when it is thinking, it will be possible to develop a mind-reading interface for robots to directly read the human mind.

1.6 Conclusion

Biologically inspired robotics is an emerging and fast growing area. It is an interdisciplinary subject that encompasses biology and engineering areas, including mechanical, electronic, control, and computer engineering Biologically inspired robotics covers the topics of robot design, sensors, actuators, control systems/algorithms, etc. Tremendous research efforts are being made in biologically inspired robotics worldwide. This book presents a collection of works on the latest developments in related topics in this ever-growing area The purpose is to help our readers gain insight and understanding of the technology and principles in this area.

References

http://www.wikipedia. org/wiki/Humanoid_robot. Last modified August 7, 2011.

Arimoto, S., Kawamura, S., Miyazaki, F., and Tamaki, S. 1985. Learning control theory for dynamic systems. Proceedings of the 24th IEEE International Conference on Decision and Control, 24(1): 1375-1380.

Arimoto, S., Naniwa, T., and Suzuki, H. 1990. Robustness of P-type learning control with a forgetting factor for robot motions. Proceedings of the 29th IEEE Conference on Decision and Control, Honolulu, HI, December 5-7, 1990.

Beer, R. D., 2009. Biologically inspired robotics. Scholarpedia. http://www. scholarpe-dia.org/article/Biologically_inspired_robotics.

Brooks, R.A. 1987. A hardware retargetable distributed layered architecture for mobile robot control. Paper read at the IEEE International Conference on Robotics and Automation, Raleigh, NC, March 31-April 3, 1987.

Fukuda, T. and Kawauchi, Y. 1990. Cellular robotic system (CEBOT) as one of the realization of self-organizing intelligent universal manipulator. Paper read at the IEEE International Conference on Robotics and Automation, Cincinnati, OH, May 13-18, 1990.

Hirai, K 1997 Current and future perpectives of Honda humanoid robot Proceedings of IEEE/RSJ Interntional Conference on Intelligent Robots and Systems, 2: 500-509.

Hirose, S. and Yamda, H. 2009. Snake-like robots. IEEE Robotics and Automation Magazine, 167(1): 88-98.

Huang, Q. Peng, Z., Zhang, W., Zhang, L. and Li, K. 2005. Design of humanoid complicated dynamic motion based on human motion capture Proceedings of IEEE/ RSJ International Conference on Intelligent Robots and Systems, pp 3536-3541, Edmonton, Canada, August 2-5, 2005

Kaneda, Y., Kamamichi, N., Yamakita, Y., Asaka, K., and Luo, Z W 2003 Control of linear artificial muscle actuator using IPMC. Paper read at the SICE Annual Conference, Fukui, Japan, August 4-6, 2003.

Kaneko, K., Kanehiro, F., Kajita, S., Hirukawa, H., Kawasaki, T., Hirata, M., Akachi, K., and Isozumi, T. 2004. Humanoid Robot HRP-2. Proceedings of2004IEEE/RSJ International Conference on Robotics and Automation, New Orleans, Louisiana, April 2004

Liu, J., Dukes, I., and Hu, H. 2005. Novel mechatronics design for a robotic fish. Paper read at the IEEE/RSJ International Conference on Intelligent Robots and Systems, Edmonton,Canada, August 2-6, 2005

Liu, Z., Ma, S., Li, B., and Wang, Y.C. 2006. 3D locomotion of a snake-like robot controlled by cyclic inhibitory CPG model. Paper read at the IEEE/RSJ International Conference on Intelligent Robots and Systems, Beijing, October 9-15, 2006.

Movellan, J. R., Tanaka, F., Fortenberry, B., and Aisaka, K. 2005. The RUBI-QRIO Project: Origins, principles, and first steps. Proceedings of the 4th International Conferneces on Development and Learning, pp. 80-86, Osaka, Japan, July 19-21, 2005

Oguro, K., Kawami, Y., and Takenaka, H 1992 Bending of an ion-conducting polymer film-electrode composite by an electric stimulus at low voltage. Journal of the Micromachine Society, 5: 27-30 (in Japanese)

Parker, J. K., Khoogar, A. R., and Goldberg, D. E. 1989. Inverse kinematics of redundant robots using genetic algorithms Paper read at the IEEE International Conference on Robotics and Automation, Scottsdale, AZ, May 14-19, 1989.

Shibata, T 2004 An overview of human interactive robot for psychological enrichment Proceedings of the IEEE, 92(11): 1794-1758

Walter, W. G. 1963. The Living Brain, 2nd ed. New York: W. W. Norton and Company.

Wang, X., Matsushita, T., Sagara, S., Katoh, R., and Yamashita, T. 1992. Two approaches to force control by rubbertuator-driven mechanism applications of IMC and SMC Paper read at the International Conference on Industrial Electronics, Control, Instrumentation and Automation, San Diego, CA, November 9-13, 1992