Automation, Robotics, and Control

The Automation, Robotics, and Process Control group is concerned with the development of technologies for controlling the dynamics of complex systems for a variety of applications.

Faculty in this research area include: Sean Andersson, John Baillieul, Calin Belta, Christos Cassandras, David CastaƱon, Pierre Dupont, Michael Gevelber and Hua Wang.

Automation:

automationState-of-the-art automation solutions are developed for a wide range of industries, ranging from sub-micron, high precision machinery for optoelectronics and semiconductor manufacture, to high-speed assembly of consumer goods. Our data-driven systems facilitate flexibility, self diagnostics and minimal change over time. A representative project in this area is the development of an automated fiber-optic coil winder for fiber-optic gyroscope production. This computer-controlled machine reduces cycle time by an order of magnitude, actively controls winding tension, winds several patterns with zero change-over time and includes a vision-based error detection and correction system for increased quality and reliability.

Process control:

Process ControlImproving materials processing capabilities is of fundamental importance to meeting the manufacturing requirements for the next generation of materials. However, processing problems are increasingly more difficult as we seek to manufacture new materials, achieve greater control over material microstructure, and meet stringent performance requirements.

Research projects focus on developing and applying a controls-based approach to enhancing materials processing capabilities which includes an integrated effort of physical modeling, sensor development, system design and control development. Design of both the system and control structure needs to be conducted with an understanding of the dominant process physics. Research projects, typically conducted with industry partners, span a range of important applications areas including enhanced materials processing for opto-electronic applications, advanced engines, power systems and cutting tools.

Robotics:

Specific applications now under development include integration of mechanical and control system design for industrial robot manipulators, adaptive tuning of adjustable vibration absorbers for aircraft and space structures, active and/or passive methodologies for introducing motion confinement characteristics in repetitive structures, distributed impact damping as a means of attenuating vibrations of machinery and structures, friction modeling and compensation applied to precision machining, disk head positioning, teleoperation, and robotic manipulation, as well as biologically-inspired control of walking robots.

Intelligent machines and information based control:

A natural extension of computer-based control is the introduction of modern wired and wireless networking technologies to integrate spatially distributed sensors, actuators, and other devices to close control loops. Research is focused on the use of wireless networking technologies (IEEE 802.11b, Bluetooth, etc.) for both hard real-time (teleoperated robot control, robotic surgery, etc.) and soft real-time (supervisory control of mobile robots) applications. Also, work is aimed at understanding the fundamental information processing (data-rate) requirements for controlling complex physical systems. Data compression techniques which optimize closed-loop control performance are being developed for communication links between sensors and actuators that are subject to network congestion or other data-rate limiting factors.

David CastaƱon