George A. Bekey, Ph.D. - US grants

A neural network model for task allocation among multiple mobile robots is proposed. The goal of the system is to allocated tasks adaptively while optimizing an overall performance index and avoiding conflicts among the robots. Robots will be allocated to work locations according to the activation levels of corresponding units in the network. Adaptiveness to environmental or internal system changes will be achieved through a process of constraint propagation and relaxation. It is proposed to resolve potential collisions among the moving robots by analyzing a global path planning problem followed by a local path replanning problem. Structural changes in the connectionist network will provide the information for conflict resolution in replanning to avoid collisions. The usefulness of the framework for controlling mobile robots will be demonstrated by simulation on a TI Explorer Lisp machine, using an object-oriented simulation language being developed in PI's laboratory.

Because most products can be made more cheaply outside its boundaries the United States is slowly losing its manufacturing base. Companies are now recognizing that they must improve their productivity if they are to maintain a market share in world markets. Research addressing manufacturing systems that are cost effective is necessary to maintain our productivity and position in world trade. Companies are now recognizing and supporting research on manufacturing. Because of this increasing interest by companies, an Industry/University Cooperative Research Center on Manufacturing and Automation now appears to be both feasible and viable. This five.year continuing grant is initiating an Industry/University Cooperative Research Center for Manufacturing and Automation at the University of Southern California and the University of California.Los Angeles in collaboration with the California Institute of Technology, the Universities of California Irvine and Santa Barbara, and Arizona State University. The initial research program is addressing: 1) An Intelligent System for Automatic Generation of Manufacturing Process Plans; 2) Geometric Modeling and its Applications in Design and Manufacturing Automation; 3) Automatic Generation of Optimal Cutting Conditions for NC Machines in CAD/CAM; 4) Multi.Robot Assembly Process Planning and Scheduling; 5) Design for Automatic Assembly and Maintenance; 6) Intelligent Simulation Systems for Automated Factories; 7) Sensor Based Robot Systems for Automated Manufacturing in a Vacuum Environment; 8) Dynamic Scheduling for Automated Manufacturing. Four companies have joined the Center and another four have said they will join at the beginning of the year when new budgets are available. Industrial support will provide 40% of the Center's $1,000,000 initial budget and should increase to 60% during the first year. The State of California has budgeted $200,000/year to augment the Center's research. The Co.Principal Investigators and all his colleagues involved in the Center are noted experts in there fields and have the expertise and industrial relationships to operate this Center. This Center is being cost shared with the Division of Design, Manufacturing, and Computer.Integrated Engineering. The Program Manager recommends the University of Southern California (and five associated universities) be awarded a five.year continuing grant to be funded at $400,000 per year. This includes $100,000 per year from the Division of Design, Manufacturing, and Computer.Integrated Engineering. . F F. Near the end of each 12.month period, the Program Manager and/or the Head of The Division of Cross.Disciplinary Research will review the progress of the Center on a number of renewal criteria, including the following: (1) extent to which university/industry interaction and collaboration is developing; (2) extent to which the support base for the center is expanding; (3) extent to which a robust research program is developing; (4) As industrial support increases, all participating universities are to achieve a fair share as determined by the.j . industrial advw is satisfactory, the Program ll recommend support of the next period of this tinuing grant.

A 32 - node hypercube multi computer for researchers at the University of Southern California for research in the Department of Computer Science. This equipment is provided under the Instrumentation Grants for Research in Computer and Information Science and Engineering program. The research for which the equipment is to be used will be in the areas of: Intelligent robotics and control Medium-to-large grain distributed artificial intelligence (DAI) Computational neurobiology and neural computing Computer vision and image understanding Parallel computer architectures for AI

This proposal is to develop a novel approach to the eventual design of a hand prosthesis. The PIs will first develop a formal model of grasping behavior based upon experiments on humans. They will then formulate a minimal set of control inputs that is necessary to control grasping movements. Finally, both neural network and knowledge-based approaches will be investigated as possible methods of implementing intelligent control mechanisms for an already developed hand prosthesis. The proposed work should lead to better ways of controlling artificial hands; it also may have an impact on the design of robotic manipulators.

Setup and fixturing are amongst the least automated of all the production activities for mechanical components and assemblies, and yet they contribute significantly to overall factors such as cost and lead time. This project seeks to automate the entire fixturing process, from the design of fixtures to their physical realization as assemblies of modular components. Theory and technology from the fields of kinematics and dynamics, geometric modeling, robotics, and artificial intelligence will be deployed through a joint effort involving four principal investigators and their students in two institutions, plus an industrial collaborator. Multiple tasks will be considered-for example, NC (numerically controlled) machining, metal-deposition processes, assembly, and inspection. Task descriptions, plus solid models of parts, assemblies and modular fixturing components will be used to compute fixture configurations appropriate for the tasks, and to generate plans for assembling the fixtures automatically. In addition, issues of design for fixturability will be addressed, and feedback will be provided to designers about the fixturing consequences of their decisions. The main objectives of the project are to advance the current understanding of representations and algorithms for automating the various steps of the overall fixturing process (e.g., the synthesis of fixture configurations), and to integrate subproblem solutions into a prototype, experimental system that demonstrates the feasibility of the approach. An industrial collaborator will provide a test bed for experimental investigations, information on current industrial practices, and realistic tasks and products.

In this research, neural network and schemalevel models of visual- motor conditional learning in monkeys will be extended to cover a wider range of reaching and grasping behaviors. These models will then be used as a basis for robot learning, including such paradigms as reinforcement learning, staged learning, focus of attention, and learning by showing. The resulting neural net strategies will then be used to enable a robot to learn to use visual and tactile input to grasp arbitrary objects, to construct assemblies of blocks, and to coordinate the motion of two arms. Existing tools for neural network simulation will be interfaced with software for robot sensing and control.

9872488
Rickel

A new three-year project is proposed, aimed at significantly expanding the VFTS through extensive evaluation and dissemination. After preparing the software for large-scale dissemination, and instrumenting it to gather evaluation data, controlled testing and evaluation will begin at four different universities: USC, San Jose State University, Tennessee State University, and the University of Virginia. Use within each university as well as the use of virtual teams that span universities will be evaluated. Next, they intend to promote the use of the VFRS to other engineering and business schools through conference workshops, promotional materials, and publications. They will also introduce the VFTS to Los Angeles area high school students as a way of promoting engineering careers, while encouraging other universities to take similar steps in their communities. Because the VFRS is Webbased and network-enabled, they will not need to produce and sell CD-ROMS for its distribution, nor will we need to coordinate software purchases. Thus, the continued dissemination and use of the VFTS beyond the three-year period of this proposal can proceed without additional funding, and with a minimum of administrative oversight.

To achieve a national impact on the education of future manufacturing engineers, a large, interdisciplinary team is crucial. The team includes computer scientists with extensive experience developing web-based educational software, engineering professors at four different universities with extensive experience teaching factory management concepts and methods, and experts in educational theory and evaluation methods. They have assembled an advisory board from industry to ensure that our educational goals match industry needs. Finally, they have individuals from industry, universities, and high schools that are committed to the broad dissemination of the VFTS and its associated approach to education.

The Multihop wireless capabilities will enable communication and coordination among autonomous nodes in unplanned environments and configurations. At the same time wireless channels present challenges of dynamic operating conditions, power constraints for autonomously-powered nodes, and complicating interactions between high level behavior and lower level channel characteristics (e.g. , increased synchronized communication will significantly degrade channel characteristics). The major goal of the research is the development, testing and characterization of algorithms for scalable, application driven, wireless network services using a heterogeneous collection of communicating mobile nodes. Some of these nodes will be autonomous (robots) in that their movements will not be human controlled. The others will be portable thus, making them dependent on humans for transportation. While the focus of the work is on the mobile nodes, include immobile computers on the network as well. It emphasize that most (though not all)
of the mobile nodes will have modest sensing, computational and communication resources. The collection as a whole is an example of a sensor network. The chief scientific motivation behind the work is the design of robust, efficient and scalable algorithms. The researchers hypothesize that distributed algorithms that rely on local interactions have many compelling characteristics, resulting in these properties. There is significant overlap between the problems of coordinating the autonomous mobile nodes that carry some of the sensors and the algorithms that direct the f ow of information from source(s) to sink(s) in the network. Both setsof algorithms will be carefully designed to improve robustness, effciency and scalability.

0075691
Bekey

This workshop will attempt to identify fundamental and applied research issues associated with very large construction and manufacturing systems in space and/or on the moon. Such issues will arise in connection with the assembly of solar power satellite systems (which may require the use of hundreds of autonomous, free-flying robots) and with the manufacture of photo-voltaic systems on the moon. The proposed workshop will concentrate on the system issues, including control, intelligence, cooperation, and self-replication. Partial support for the workshop has been obtained from NASA Headquarters.
***