Biorobotic approaches to the study of motor systems

Current Opinion in Neurobiology

Volumn 8, Issue 6, 1998, Pages 777-782

  Beer, Randall D ^a   Chiel, Hillel J ^b   Quinn, Roger D ^a   Ritzmann, Roy E ^c  

^a Case Western Reserve University   (United States)

^b CASE WESTERN RESERVE UNIVERSITY SCHOOL OF MEDICINE   (United States)

^c CASE WESTERN RESERVE UNIVERSITY   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

HUMAN; LIMB MOVEMENT; LOCOMOTION; MOTONEURON; NONHUMAN; PRIORITY JOURNAL; REVIEW; ROBOTICS;

EID: 0031674216     PISSN: 09594388     EISSN: None     Source Type: Journal    
DOI: 10.1016/S0959-4388(98)80121-9     Document Type: Article

References (54)

1. Beer R. A dynamical systems perspective on agent-environment interaction. Artif Intell. 72:1995;173-215.
2. Ashby WR. Design for a Brain. 1952;Chapman and Hall, London.
3. Chiel HJ, Beer RD. The brain has a body: adaptive behavior emerges from interactions of nervous system, body and environment. of outstanding interest Trends Neurosci. 20:1997;553-557 Recent experimental evidence that the physical properties of bodies and environments are crucial for understanding the operation of neural circuits is reviewed. Recent work in the combined modeling of an animal's neural circuitry, peripheral biomechanics and environment (computational neuroethology) is also reviewed.
4. Beer RD. Intelligence as Adaptive Behavior: An Experiment in Computational Neuroethology. 1990;Academic Press, Boston, Massachusetts.
5. Cliff D. Computational neuroethology: a provisional manifesto. Meyer J-A, Wilson SW. From Animals to Animals: Proceedings of the First International Conference on Simulation of Adaptive Behavior. 1991;29-39 MIT Press, Cambridge, Massachusetts.
6. Beer RD, Ritzmann RE, McKenna T. Biological Neural Networks in Invertebrate Neuroethology and Robotics. 1993;Academic Press, Boston, Massachusetts.
7. Miki N, Shimoyama I. Analysis of the flight performance of small magnetic rotating wings for use in microrobots. Proceedings of the IEEE International Conference on Robotics and Automation. 1998;3065-3070. 1998 May 16-20. Leuven, Belgium.
8. Saito F, Fukuda T, Arai F. Swing and locomotion control for a two-link brachiation robot. IEEE Control Systems Magazine. 14:1994;5-12.
9. Hirose S. Biologically Inspired Robots: Snake-Like Locomotors and Manipulators. 1993;Oxford University Press, Oxford.
10. of special interest Vaidyanathan R, Chiel HJ, Quinn RD. A hydrostatic robot for marine applications. Proceedings of the Eleventh VPI & SU Symposium on Structural Dynamics and Control. 1997;543-551 Virginia Polytechnic Institute and State University, Describes the design of a worm-like hydrostatic robot consisting of three fluid-filled bladders connected by sets of four shape memory alloy springs. By using the appropriate differential activation of the four springs, the robot was capable of undulatory locomotion underwater and turning around obstacles.
11. Raibert MH, Hodgins JK. Legged robots. Beer RD, Ritzmann RE, McKenna T. Biological Neural Networks in Invertebrate Neuroethology and Robotics. 1993;319-354 Academic Press, Boston, Massachusetts.
12. Pratt J, Dilworth P, Pratt G. Virtual model control of a bipedal walking robot. of special interest Proceedings of the IEEE International Conference on Robotics and Automation. 1997;193-198 Describes an intuitive approach to controller design termed 'virtual model control', in which control is achieved by simulating the effects of virtual mechanical components attached to a physical robot. The authors used this approach to control support and walking in a planar bipedal robot with two actuated DOF per leg.
13. Yamaguchi J, Takanishi A. Development of a biped walking robot having antagonistic driven joints using nonlinear spring mechanism. 1997 April 20-25. Albuquerque, New Mexico Proceedings of the IEEE International Conference on Robotics and Automation. 1997;185-192.
14. Hirai K, Hirose M, Haikawa Y, Takenaka T. The development of Honda humanoid robot. of special interest Proceedings of the IEEE International Conference on Robotics and Automation. 1998;1321-1326 Although only loosely inspired by biology, the bipedal robot developed by these authors is one of the most sophisticated free-walking humanoid robots in existence, with six DOF per leg and seven DOF per arm. It can walk forward, backward, or sideways, walk up or down stairs, step over obstacles, turn in any direction, and maintain its posture on uneven terrain.
15. Donner M. Real-Time Control of Walking. 1987;Birkhauser Boston, Cambridge, Massachusetts.
16. Brooks RA. A robot that walks: emergent behaviors from a carefully evolved network. Neural Computation. 1:1989;253-262.
17. Pearson KG, Franklin R. Characteristics of leg movements and patterns of coordination in locusts walking on rough terrain. Int J Robotics Res. 3:1984;101-112.
18. Delcomyn F, Nelson ME, Cacatre-Zilgien JH. Sense organs of insect legs and the selection of sensors for agile walking robots. Int J Robotics Res. 15:1996;113-127.
19. Cruse H, Brunn DE, Bartling C, Dean J, Dreifert M, Kindermann T, Schmitz J. Walking: a complex behavior controlled by simple networks. Adaptive Behavior. 3:1995;385-418.
20. Pfeiffer F, Eltze J, Weidemann HJ. The TUM walking machine. Intl J Automation Soft Computing. 1:1995;307-323.
21. Cruse H. What mechanisms coordinate leg movement in walking arthropods. Trends Neurosci. 13:1990;15-21.
22. Cruse H, Kindermann T, Schumm M, Dean J, Schmitz J. Walknet: a biologically inspired network to control six-legged walking. Neural Networks. 1998;. in press.
23. Beer RD, Quinn RD, Chiel HJ, Ritzmann RE. Biologically inspired approaches to robotics. of outstanding interest Commun Associ Comput Machinery. 40:1997;30-38 Reviews a series of insect-like legged robots whose control is based on experimental studies of leg coordination and local leg reflexes in the cockroach, walking stick, and locust. The most sophisticated robot can walk with a range of insect gaits at speeds up to 14 cm/s and negotiate irregular, slatted and compliant surfaces. The robotic application of dynamical neural circuits for walking that were evolved using genetic algorithms is also briefly described.
24. of outstanding interest Bachman RJ, Nelson GM, Flannigan WC, Quinn RD, Watson JT, Tryba AK, Ritzmann RE. Construction of a cockroach-like hexapod robot. Proceedings of the Eleventh VPI & SU Symposium on Structural Dynamics and Control. 1997;647-654 Virginia Polytechnic Institute and State University, Blacksburg, Virginia, Describes the mechanical design of a new cockroach-like hexapod robot closely modeled after Blaberus discoidalis. The robot was constructed from machined aluminum, weighs 30 pounds, and is 17 times the size of the insect. Based on kinematic studies of Blaberus, the robot has five DOF in each front leg, four DOF in each middle leg, and three DOF in each rear leg. These joints are actuated by pneumatic cylinders whose valves are driven with 50 Hz pulse width modulation. Potentiometers on each joint provide joint angle feedback.
25. Watson JT, Ritzmann RE. Leg kinematics and muscle activity during treadmill running in the cockroach, Blaberus discoidalis: I. Slow running. J Comp Physiol [A]. 182:1997;11-22.
26. Watson JT, Ritzmann RE. Leg kinematics and muscle activity during treadmill walking in the cockroach, Blaberus discoidalis: II. Fast running. J Comp Physiol [A]. 182:1997;23-33.
27. Nelson GM, Quinn RD. Posture control of a cockroach-like robot. of outstanding interest Proceedings of the IEEE International Conference on Robotics and Automation. 1998;157-162 Describes a postural controller for the cockroach-like hexapod robot described in [24] based on a virtual model control scheme that uses a single virtual leg model of the robot's mechanics to encourage equal load distribution and minimize joint torques in the supporting legs. With this control scheme, the robot can adjust its posture to commanded body positions and orientations, smoothly recover from substantial postural perturbations, and lift a 30 pound payload.
28. Triantafyllou MS, Triantafyllou GS. An efficient swimming machine. Sci Am. 1995, March;64-70.
29. Kumph JM, Triantafyllou MS. A fast-starting and maneuvering vehicle, the RoboPike. of outstanding interest Proceedings of the Symposium on Seawater Drag Reduction. 1998; The authors describe the design of a free-swimming robotic fish modeled closely on the chain pickerel Esox niger. The main body of the robot is made of fiberglass spiral wound spring covered by a skin of latex, lycra and steel mesh. The body is 81 cm long and has two main articulations controlled by servo motors. Additional servo motors control caudal and pectoral fins. To date, open-loop control has been used to produce fast C-starts and forward swimming at speeds up to 1 m/s.
30. Mojarrad M, Shahinpoor M. Biomimetic robotic propulsion using polymeric artificial muscles. Proceedings of the IEEE International Conference on Robotics and Automation. 1997;2152-2157 A polyelectrolyte ion exchange membrane-metal composite muscle-like actuator was used as a caudal fin for a robotic swimmer. This material bends as a result of ion redistribution in response to an imposed electric field. By driving the fin with an oscillatory voltage signal, a traveling bend was produced that propagated along the length of the actuator, shedding vortices that propel the robotic fish.
31. Kato N, Inaba T. Guidance and control of fish robot with apparatus of pectoral fin motion. Proceedings of the IEEE International Conference on Robotics and Automation. 1998;446-451. 1998 May 16-20. Leuven, Belgium.
32. Franceschini N, Pichon JM, Blanes C. From insect vision to robot vision. Philos Trans R Soc Lond [Biol]. 337:1992;283-294.
33. Hoshino K, Mura F, Morii H, Suematsu K, Shimoyama I. A small-sized panoramic scanning visual sensor inspired by the fly's compound eye. of special interest Proceedings of the IEEE International Conference on Robotics and Automation. 1998;1641-1646 Describes a 30 mm compound eye with 2 × 30 photodiode receptors based on a fly eye. A piezoelectric actuator rotates the eye about its center of curvature. The authors demonstrate that active scanning enhances the perception of visual motion for such a low resolution sensor. These compound eyes have been used for obstacle avoidance, tracking and pursuit of a visual target in a small mobile robot.
34. Chen C-T, Quinn RD, Ritzmann RE. A crash avoidance system based upon the cockroach escape response circuit. of special interest Proceedings of the IEEE International Conference on Robotics and Automation. 1997;2007-2012 The authors describe a neural network crash avoidance system for wheeled vehicles based on the neural circuitry controlling the cockroach escape response. The network receives proximity, wheel speed and steering angle inputs and outputs vehicle acceleration and steering angle. The network was trained using backpropagation to produce successful avoidance responses.
35. Grasso F, Consi T, Mountain D, Atema J. Locating odor sources in turbulence with a lobster inspired robot. Maes P, Mataric M, Meyer J-A, Pollack J, Wilson SW. From Animals to Animals: Proceedings of the Fourth International Conference on Simulation of Adaptive Behavior. 1996;104-112 MIT Press, Cambridge, Massachusetts.
36. Rucci M, Wray J, Tononi G, Edelman GM. A robotic system emulating the adaptive orienting behavior of the barn owl. of outstanding interest Proceedings of the IEEE International Conference on Robotics and Automation. 1997;443-448 Using a robotic head with a camera and two lateral microphones surrounded by an array of lights and buzzers, the authors explore the use of visual information to calibrate auditory orientation in a model of plasticity in the inferior colliculus and optic tectum of the barn owl. The robotic head develops spatial registration between auditory and visual maps in the model optic tectum. In addition, the robot can appropriately adapt these maps after the visual image is shifted through prismatic goggles and after the removal of these goggles.
37. Morse TM, Ferrée TC, Lockery SR. Robust spatial navigation in a robot inspired by chemotaxis in C. elegans. of outstanding interest Adaptive Behavior. 6:1998;393-410 Describes a mobile robot that navigates using phototaxis, based on studies of chemotaxis in Caenorhabditis elegans. The robot moves forward at a constant velocity with a controllable steering angle. A single photocell at the front of the robot senses light intensity. Linear dynamic neural networks based loosely on the chemotaxis circuitry of C. elegans were optimized for phototaxis in simulation and then transferred to the robot. The trajectories of the robot closely matched those observed in C. elegans and were robust to perturbations in instantaneous speed and steering angle. An analysis of the optimized networks provided insight into its operator.
38. Webb B. Using robots to model animals: a cricket test. Robotics Autonomous Systems. 16:1995;117-134.
39. of outstanding interest Lund HH, Webb B, Hallam J. A robot attracted to the cricket species Gryllus bimaculatus. Husbands P, Harvey I. Fourth European Conference on Artificial Life. 1997;246-255 MIT Press, Cambridge, Massachusetts, A small wheeled robot was used to test a model of cricket phonotaxis. The robot includes microphones representing left/right pairs of tympani and spindles and an electronic model of the cricket auditory system. The outputs of this auditory circuit are fed into leaky integrators and whichever side reaches threshold first causes a small turn in that direction. The robot was exposed to a recorded calling song of male Gryllus bimaculatus. The authors demonstrate that this simple model can account for both phonotaxis and frequency selectivity due to the frequency dependence of ear directionality.
40. Walker VA, Peremans H, Hallam JCT. Good vibrations: exploiting reflector motion to partition an acoustic environment. of outstanding interest Robotics Autonomous Systems. 1998; A robot is used to test a model of prey localization via binaural echolocation in insectivorous bats that emit long, constant frequency calls. The robot is wheeled and contains an ultrasonic transmitter and a pair of ultrasound receivers. The robot operates by making binaural comparisons of the spectral sideband energy in echoes from periodically moving targets, steering so as to cancel interaural differences. Selectivity is achieved by filtering out all sideband energy except that from the particular sidebands specific to the oscillating targets of interest (e.g. insect wingbeats), allowing robust target localization even in acoustically cluttered environments.
41. Walker VA, Peremans H, Hallam JCT. One tone, two ears, three dimensions: a robotic investigation of pinnae movements used by rhinolophid and hipposiderid bats. of outstanding interest J Acoust Soc Am. 104:1998;569-579 The authors extend the work described in [40] to investigate the consequences of pinnae movements in bats by allowing the robot's ultrasound receivers to move via servomotors. They find that such movements have two benefits: first, they produce different viewing perspectives over which interaural intensity disparities can be sampled, and second, they produce amplitude modulations of returning echoes that vary systematically with target elevation.
42. Mussa-Ivaldi FA, Giszter SF. Vector field approximation: a computational paradigm for motor control and learning. Biol Cybern. 67:1992;491-500.
43. Williamson MM. Postural primitives: interactive behavior for a humanoid robot arm. Maes P, Mataric M, Meyer J-A, Pollack J, Wilson SW. From Animals to Animats: Proceedings of the Fourth International Conference on Simulation of Adaptive Behavior. 1996;124-131 MIT Press, Cambridge, Massachusetts.
44. Brooks RA, Stein LA. Building brains for bodies. Autonomous Robots. 1:1994;7-25.
45. Sternad D, Schaal S. Segmentation of endpoint trajectories does not imply segmented control. of outstanding interest Exp Brain Res. 1998; This paper examines two hypotheses regarding the motion primitives underlying movement segmentation in three-dimensional drawing movements by humans: firstly, a stroke-based primitive suggested by abrupt changes in the velocity gain factor of a 2 3 power law relationship between curvature of the drawing path and angular velocity of the endpoint trajectory; and secondly, a segmentation hypothesis inferred from the piecewise planarity of endpoint trajectories. Using a seven DOF anthropomorphic arm, the authors demonstrate that continuous movements at the level of the biomechanics of the arm can account for the observed segmentation of endpoint trajectories due to nonlinearities in arm kinematics.
46. Hannaford B, Winters JM, Chou C-P, Marbot P-H. The anthroform biorobotic arm: a system for the study of spinal circuits. Ann Biomed Eng. 23:1995;399-408.
47. Chou C-P, Hannaford B. Study of human forearm posture maintenance with a physiologically based robotic arm and spinal level neural controller. of outstanding interest Biol Cybern. 76:1997;285-298 A robotic model of the human elbow is used to explore the maintenance of human forearm posture by spinal circuitry. The robotic elbow includes two McKibben braided pneumatic actuators whose attachment points mimic those of human elbow flexor and extensor muscles. The model also includes model Golgi tendon organs and muscle spindles. The authors explore open loop stiffness control via co-contraction, closed loop stiffness control with la or lb afferent feedback, and posture maintenance using a model of spinal reflex circuitry. They also demonstrate how gamma dynamic excitation is essential to produce velocity feedback of the la signal, which increases closed-loop damping, and how alpha motoneuron activation compensates for the lowpass filtering properties of muscle by acting as a feedforward phase-lead controller.
48. Takanobu H, Yajima T, Nakazawa M, Takanishi A, Ohtsuki K, Ohnishi M. Quantification of masticatory efficiency with a mastication robot. of special interest Proceedings of the IEEE International Conference on Robotics and Automation. 1998;1635-1640 The authors use a robotic model of the human jaw to quantify the mastication efficiency of different mandibular motions. The robot consists of an epoxy resin human skull replica with a three DOF jaw actuated by nine motors with force sensors that model the major jaw muscles. Using a small ball-like cookie, they compared the efficiency of two different chewing motions and found a grinding motion to be more efficient than a clenching motion.
49. Takanishi A, Hirano S, Sato K. Development of an anthropomorphic head-eye system for a humanoid robot. of special interest Proceedings of the IEEE International Conference on Robotics and Automation. 1998;1308-1314.
50. Almássy N, Edelman GM, Sporns O. Behavioral constraints in the development of neuronal properties: a cortical model embedded in a real-world device. Cereb Cortex. 8:1998;346-361 Using a wheeled robot equipped with a camera and a 'taste' sensor (based on electrical conductivity), the authors explore the development of selective and invariant responses of cortical neurons to complex visual stimuli. When embedded in an environment containing conductive and nonconductive objects with different visual patterns, the robot develops an attraction to visual patterns associated with nonconductive objects and an aversion to visual patterns associated with conductive objects. They found that the robot's movement was essential to the development of selective and translation invariant cortical responses, and that these responses depended on the physical design of the robot and the relative frequency of the various objects in the environment.
51. Kovacs GTA. Micromachined Transducers Sourcebook. 1998;WCB/McGraw-Hill, Boston, Massachusetts.
52. Caldwell DG, Medrano-Cerda GA, Bowler CJ. Investigation of bipedal robot locomotion using pneumatic muscle actuators. of special interest Proceedings of the IEEE International Conference on Robotics and Automation. 1997;799-804 Describes the use of McKibben pneumatic actuators as flexors and extensors for the hip and knee joints of a pair of robotic legs. Standing and striding experiments demonstrate that such actuators are promising artificial muscles that are fast, light, powerful, and compliant.
53. Kornbluh R, Pelrine R, Eckerle J, Joseph J. Electrorestrictive polymer artificial muscle actuators. Proceedings of the IEEE International Conference on Robotics and Automation. 1998;2147-2154. 1998 May 16-20. Leuven, Belgium.
54. Schwind WJ, Koditschek DE. Characterization of monoped equilibrium gaits. Proceedings of the IEEE International Conference on Robotics and Automation. 1997;1986-1992. 1997 April 20-25 Albuquerque, New Mexico.