Patrick McDowell
Download 0.83 Mb.
|
- Navigate this page:
- Summary of Current Work
- Practical Applications
- Stealthy Robotic Squirrel
- Fire Fighting Robotic Ants
| Biological
As discussed earlier, the basic proportions of the robot skeleton were derived from measurements, anatomical drawings and photographs. Figure 14 below is photograph of a cat skeleton. 
Figure 14. Cat skeleton. Note the likeness of the proportions to the mockup in figures 1 through 3 and the original prototype in figure 6. Summary of Current Work The mock-up, and the two prototypes Mickey and Stubby, have been invaluable research tools. With these prototypes, experience has been gained in several different areas, including: Materials selection, handling, and fabrication. Actuator selection and integration. Electronics interfacing and control. Sensor fusion. Algorithm design, gait creation, selection, and control. Systems integration. With Stubby, the project has a walking prototype from which to build on. Further work with Stubby includes integration of sensor inputs, adaptation of walking movements, and adaptation of the timing of the walking movements. These technologies and procedures will carry over into the development of the next prototype and finally to the finished product, the LSU Robo-Tiger. Together, the prototypes have shown that the mechanical and software design processes go hand in hand during research and development. Static and dynamic stability as well as power to weight ratio are also extremely important. Taking cues from biology at the functional and conceptual level, at the very least, has been a great help. The architecture of the control system is an excellent example of this. Relevance Although the Robo-Tiger will be primary used at school athletic functions, the technologies integrated, applied and created will have exciting and cutting edge applications in a wide variety of real world situations. One area of special interest is adaptive control. By nature, the environment of the football field is unstructured. To look and act in a realistic manner, the Robo-Tiger will have to adapt to a variety of different field conditions. For example, walking a set pattern in the lab will probably not work on the field where traction conditions are completely different and navigation around obstacles may be required. Movements will be able to be choreographed to some degree, but not in strict sense. One of the main goals of the design team is not to have to carry the tiger off the field. To make the tiger act in a realistic manner is going to require integration of low level mechanical control, adaptive control, and at the highest level, sensor/situation driven adaptation of behaviors. The subsumption based control architecture that Stubby uses is the ideal foundation for building the functionality of the robot. Because it is a layered, hierarchical approach, its capabilities can be extended one piece at a time. Complexity it achieved by adding layers without disturbing the layers below. New layers are not added until the existing layers are working well. Thus, behaviors are arranged in hierarchies, with the simplest levels at the bottom and the more complex activites on the top. The lower layers continue to operate, unaware of the layers above them. Another area of interest is in actuator design. Currently most robots use mechanical actuators that are driven electrically, pneumatically, or with hydraulics. Each of these methodologies has it advantages and disadvantages, but biological muscles have a the distinct advantage in that they are quiet. A real tiger, or any animal that relied upon stealth for purposes of pursuit or evasion would not survive its muscles were whining or hissing like the mechanical muscles do. However, in the recent years significant breakthroughs have taken place in smart materials that change shape and exert force under chemical or electrical stimuli. Moreover, such changes are reversible, i.e electrical field or potential is generated when the material is deformed. The materials are collectively known as artificial muscles. Notable species of such artificial muscles include the Ionic-polymer Metal Composite (IPMC) material developed at the University of New Mexico, the carbon Nanotube material developed by the Alliedsignal Inc., Electrostrictive Polymers developed by the SRI International, Electroactive Polymers (EAP) developed at NASA JPL, and the polymer Hydrogel developed at MIT. It has been demonstrated that some of these materials possess force density and response time equal or exceeding natural muscles. Therefore they have great potential to become viable robotic actuators. Indeed, a boat propelled by a waving fin, a pair of waving wings and a full-size skeleton bicycle rider driven by IPMC muscles have been built and demonstrated at the Artificial Muscle Research Institute of University of New Mexico. NASA JPL has also demonstrated robot grippers and wipers using the EAP muscles. The reversible change of the polymer type artificial muscles also make them effective sensors, jut like natural muscles. The collection of the sensor and actuator is a highly desirable feature for feedback motion control, as it eliminates transport delay in the feedback loop, thereby increasing the bandwidth (reducing response time) and domain of stability. While the feasibility of using IPMC type of artificial muscles as robotic actuators have been demonstrated, research is needed in learning the dynamic properties and motion control of such materials before they can be used to produce fine coordinated robotic motions. The dynamics of these materials are inherently nonlinear, distributed and elusive to mathematical modeling and conventional motion control algorithms. The distributed sensory and coordinated motions also call for a hierarchical and massively parallel control system involving higher level intelligence (logical reasoning performed by the frontal lobe and the motor lobe of the cortex) for motion planning and lower level intelligence (reflexive decision and control performed by the cerebellum) for agile and stable actuation. Practical Applications Here we outline three concrete applications which can draw directly from the systems and technologies developed for the Robo-Tiger. Walking stretcher. In the unstructured environments battlefields, natural disasters, etc., medical personal struggle against time, fatigue, and harsh conditions in an attempt to save lives. Getting injured personnel to safety can be work intensive, especially in rough terrain. In order to evacuate one injured man, at least two men are required to give emergency care and transport him to where he can more closely attended to. Using technology similar to that used to create the Robo-Tiger, a walking stretcher, or robotic pack animal could be created. When a man became injured in the field during an exercise or combat, the medic would lead the walking stretcher to the man. After examination and application of the necessary medicines, the injured man would be loaded onto the stretcher, and it could be led back to the safety of the field hospital, or even make the return trip autonomously. Medical equipment such as IV’s and small monitoring equipment could be integral to the stretcher. A device such as this would free up manpower, and may well save lives by lessening exposure to risk, and by keeping the injured more comfortable during their trip back to safety. Stealthy Robotic Squirrel. A squirrel sized sensing platform using silent actuators based on the IPMC technology discussed earlier could be used to crawl into buildings and transmit data back to a base station. In order to keep the package size small, the robots would be programmed to seek out power cables in an attempt to augment their power supplies. Fire Fighting Robotic Ants Using a platform similar to that of the walking stretcher, these autonomous robots would be programmed to work in teams, like ants do, to put out fires in unstructured, hazardous environments, such as nuclear power plant, hazardous waste depots, or on ships. While various tracked and wheeled robots have been to some degree in this application, legged robots would have a distinct advantage if they were taught to climb. Environments like a burning ship or chemical factory could be attacked much more effectively and with less risk to humans because the robots would have the ability to detect, seek out, and attack the blaze. With bodies similar to fire extinguisher tanks, and four or six legs the robots would work in teams to fight fires. Extra power to operate and recharge the robots could be drawn directly from the blaze itself, via a Sterling engine or some other heat engine system. References[1] J. Zhu and W. Xaio “Intelligent control of time-varying dynamical systems using CMAC artificial neural network,” Mathematical and Computer Modeling Journal, Special Issue on neural networks, vol. 21, no. ½, 89-107, 1995. [2] J. Zhu and M. C. Mickle, “Missile autopilot design using a new linear time-varying control technique,” AIAA Journal on Guidance, Control and Dynamics, vol. 20, no. 1, 150-157, Jan. 1997. [3] J. Zhu, “Nonlinear tracking and decoupling by trajectory linearization,” a 12-hour lecture with pp. 137. Lecture note presented at NASA Marshall Space Flight Center, Huntsville, AL, Grant No. NASA/LEQSF (96-01)-01, May 1, 1998-April 30, 1999. [4] M.C Mickel and J. Zhu, “Bank-To-Turn Roll-Yaw-Pitch Autopilot Design Using Dynamic Nonlinear Inversion and PD-eigenvalue assignment,” Proceedings, 2000 American Control Conference, Chicago, IL, to appear, June 2000. [5] J. Zhu, Brad D. Banker and Charles E. Hall, “X-33 Ascent Flight Controller Design By Trajectory Linearization – A Singular Perturbational Approach,” Proceedings, AIAA Guidance, Navigation and Control Conference, Denver, CO, to appear, August, 2000.
Beckers R., Deneubourg J.L. and S. Goss (1992). Trails and U-turns in the selection of the shortest path by the ant Lasius niger. Journal of theoretical biology, 159, 397-415. Goss. S., Aron. S., Deneubourg J.L. and J.M. Pasteels (1989). Self-organized shortcuts in the Argentine ant. Naturwissenschaften 76, 579-581. Free On Line Dictionary of Computing; http://foldoc.doc.ic.ac.uk/foldoc/index.html Kevin Kelly, Out Of Control, Perseus Books, 1994 Directory: Academics -> Faculty -> pmcdowell Faculty -> Chapter 1 “Preliminaries” Reasons for Studying Concepts of Programming Languages Faculty -> Chapter 7 “Expressions and Assignment Statements” Faculty -> Deleuze and Analytic Philosophy by Faculty -> Curriculum Proposal Title of Proposal: Course Status Change Faculty -> Chapter 6: Arrays Faculty -> Chapter 5 Introduction pmcdowell -> Guide to Intelligent Systems; Second Edition; Michael Negnevitsky Faculty -> Carmen Sara Cañete Quesada Faculty -> Economic Conditions for Regional Development and for First Nations Download 0.83 Mb. Share with your friends:
|