This project investigates telemanipulation in multi-robot networks. We are addressing fundamental issues in communication, coordination, and teleoperated control of multiple agents in coordinated manipulation tasks. The combination of teleoperation and manipulation along with multi-robot coordination is, in fact, one of the novel aspects of our research. While multi-agent coordination and control problems such as swarming, flocking, and rendezvous have been studied by several researchers, much less work has gone into the teleoperated control of multi-robot networks, especially when the multi-robot network is expected to engage in tasks involving both manipulation and motion coordination. Manipulation tasks require haptic and force feedback which introduce significant stability and transparency problems with respect to communication delay, packet loss, and other communication effects.
There are many reasons to study such problems. Tasks such as remote construction, search and rescue, salvage operations, and telesurgery are too complex at the present time to be carried out by fully autonomous robots but may be carried out by networks of semi-autonomous robots, coordinated by human operators and providing sensory data back to the human operators.
An important issue that recurs throughout our investigations is the trade-off among communication, computing, and control. For example, how does one trade-off computation and control with real-time communication among master and slave robots? At one extreme, each slave robot could be equipped with sufficient on-board computation (DSP, microprocessor, memory) to operate more or less autonomously. At the other extreme, the slave robots may have little on-board computation with most computation centralized at the master. The first case greatly increases the financial cost of each robot whereas the second case greatly complicates the real-time communication and control requirements. By investigating such trade-offs we hope to develop algorithms that are not only robust with respect to communication delays, losses and other network effects, but that are also highly scalable with respect to the size of the robot network.
In addition to the algorithmic design and analysis we are building a multi-robot testbed to implement our obtained results and identify further research problems.Postdoctoral Researchers
This project is to investigate passivity based control in bipedal locomotion. In recent years, passivity based control has proven to be one of the most powerful design methodologies for the control of electromechanical systems such as robot manipulators, underwater vehicles, induction motors, automotive and aerospace systems, and others. With a few exceptions the application of these methods to walking robots and other systems with impacts has not been adequately investigated. The project will explore several extensions of bipedal locomotion in the context of passivity based hybrid nonlinear control. The project will investigate speed regulation, the use of alternate potential functions to increase the basins of attraction of stable limit cycles, the effect of control saturation and under actuation in passivity based control, and the efficiency of passivity based control methods compared to true energy optimal control. It will also investigate passivity based control of gait transitions, including starting and stopping. The goal, and the technical merit of the project, is to help solidify the foundations of the field through analysis, development of new concepts, and the design of provably correct control algorithms. Another aspect of the project is to integrate the theoretical tools of passivity based analysis and control with studies of balance and locomotion in human subjects in order to supplement the descriptive research typical in those studies with more analytical methods. The practical application of this research project is on the design of walking robots that have improved performance capabilities over existing machines. Current walking robots have limited range due to poor energy utilization and are limited in their ability to navigate rough terrain. More practical and more efficient walking machines will result once the full power of available theoretical tools is brought to bear on the analysis and design questions in this project. From a broader perspective, the applications of this research will extend beyond the design of improved walking machines. The analysis and design tools developed in this project will also contribute to a better understanding of human locomotion, which will result in applications in biomechanics and biomedicine, such as the design of improved prosthetic devices, the development of falls prevention programs for the elderly, and rehabilitation techniques. The impact of falls among the elderly in the United States alone has a yearly impact on the economy of more than ten billion dollars in medical bills and other expenses. The improved modeling and analysis tools of this project will be applied to real data obtained from human subjects in order to understand not only how aging affects balance and locomotion, but also how to develop intervention techniques to decrease the rate of falls.
This project is to develop reliable and robust control architectures and algorithms for networks of autonomous aerial and ground vehicles. The aim is to develop control laws that have low sensititivy to noisy and lossy data communication among vehicles, that are scalable in terms of number of vehicles, and that have the ability to handle discrete transitions in the network, such as formation reconfiguration, addition or loss of vehicles from the formation, etc. Applications of this owrk include undersea and planetary exploration, search and rescue, air traffic control, and control of sensor networks. Both theoretical and experimental issues are being investigated.Faculty
This award supports US-France collaboration in control systems between Mark W. Spong of the University of Illinois and Romeo Ortega of the Signal and Systems Laboratory at SUPELEC, a French center for research in electrical engineering. The objective is to investigate passive nonlinear control of networked control systems, in particular, systems involving bilateral remote operation (teleoperation) over unreliable communication networks. The problem is motivated by interest in wireless communication in embedded real time control systems and the use of the Internet as a communication medium in teleoperated and networked control systems. The research will advanced understanding of how to design and utilize mobile, intelligent robotic systems with local intelligence for communication over unreliable networks with other robots and human operators. The researchers plan to answer: How sensory, command and control information are shared over such networks and how much local intelligence is needed to guarantee performance and stability when network communications are degraded. Potential applications of their results include work in hazardous and remote environments, surveillance, search, and rescue robots, autonomous vehicles and autonomous locomotion systems, haptic devices, remote construction, and remote surgery.