Morse Named Recipient of Highest Engineering Award in Field
New Haven, Conn. -- On a nearly vertical slope in the Alps virtually inaccessible by helicopter, two mountain climbers get sick and need medical supplies badly. Normally, the sick climbers would have to wait for a rescue party to reach them. However, scientists hope that soon a small, unmanned hover craft could be dispatched from base camp that would be able to maneuver through forests and crevasses -- on its own, without remote radio control -- to bring the needed supplies to the climbers quickly.
Designing such a vehicle is the job of scientists who work in the field of control systems. One man who is planning to launch such a UAV (unmanned aerial vehicle) project is Yale electrical engineering professor A. Stephen Morse. "I'm looking forward to tackling unmanned aircraft next," Morse said recently.
Morse is not only about to embark on one of the most challenging projects in the field of control systems, he also is being awarded the 1999 Control Systems Award by the IEEE (Institute of Electrical Engineers and Engineering) for his work in the area of automatic control. The award, which is being conferred at an international conference in December 1999, is the highest in the field of control systems, and recognizes Morse for outstanding contributions to that field worldwide.
Morse, who has been a member of Yale's electrical engineering faculty since 1970, explains that automatic control systems have been a prominent part of emerging technology for decades. "Cruise controls, anti-lock brakes, pace-makers, even autopilots in airplanes are old ideas," Morse notes. "People were flying under autopilots by the end of World War II."
Perhaps the simplest example of an automatic control system is a thermostat that automatically turns a furnace on and off in order to maintain a specified temperature. Another example familiar to most people is a cruise control in a car, which regulates the amount of fuel going into the engine in order to maintain a constant speed over a variety of inclines and road surfaces.
"Both a thermostat and cruise control involve all the basic elements of an automatic control system," Morse says. "First, they employ some kind of sensor, which picks up relevant data from the overall system being controlled, whether it's temperature or speed. Next, such systems must contain a mechanism, or 'actuator,' for operating the machinery involved, such as an engine throttle. Finally, every automatic control system needs some kind of 'brain' to interpret the sensed data and adjust the actuator accordingly."
The "brain" in most cases is a computer, and that's where control systems engineers and scientists come in. "Control systems engineers devise the list of instructions, or mathematical algorithms, which tell the computer what to do," Morse explains. "Control scientists, such as myself, develop new types of algorithms."
Morse's IEEE award recognizes his contributions to two areas of automatic control system algorithms. First, he is being recognized for pioneering a new approach to "linear multivariable control synthesis." Linear algorithms involve the simplest type of algebraic equations, he says, and multivariable control systems are those which utilize many different sensors and actuators to control a piece of equipment.
A good example of a multivariable control system is an automatic pilot in an airplane. The job of an autopilot is to keep an airplane flying level, in a specified direction, and at a specified altitude. The system must sense data from the plane's rudder and wing flaps, readings from small gyroscopes that tell whether the plane is level, and altitude readings, says Morse, who did most of his work in multivariable control in the 1960s and 1970s.
The IEEE also is conferring the Control Systems Award on Morse for his contributions to "non-linear" and "adaptive control" theory, areas he has been working in since the late 1970s. Non-linear control systems employ more complex, higher-level algorithms, while adaptive control systems are automatically capable of tuning themselves to compensate for changes in performance due to aging, environmental variations, component failures, etc.
"An example of an adaptive control system is an automatic steering system, or autopilot, for a supertanker, which is able to retune itself to compensate for changes in roughness or sea currents," Morse says. "The systems that steer modern ships automatically retune themselves, and are programmed to know when to do so."
Another area in which adaptive control systems figure prominently is that of robotics. Robots are controlled by means of visual data supplied by video cameras and, as a robot moves, the positions of the cameras must remain calibrated. But cameras can lose calibration unpredictably for a number of reasons, such as excessive vibration or atmospheric changes. In an adaptive system, camera recalibration is accomplished automatically while the system is operating.
In recent years, Morse has been working extensively with his graduate students and with Yale computer scientist Gregory D. Hager on new developments in robotics. "Together, we have been able to classify those vision-guided robot positioning tasks that can be accomplished with precision even when the cameras are not perfectly calibrated," Morse said.
For earlier work with his colleagues A. Kanellakopoulos and P.V. Kokotovic, Morse received the George S. Axelby Outstanding Paper Award for development of the concept of "back-stepping," which breaks complex, nonlinear system-design procedures down into simple steps. The work grew out of experiments he performed with the help of his graduate student Arie Feuer at Yale in the late 1970s.
Recently, Morse has been working on the development of adaptive control systems that involve switching and logic -- work which overlaps that of scientists in the artificial intelligence community. "This is adaptive control with a twist," says Morse, who notes that the military has taken an interest in his work with unmanned aircraft, especially very small surveillance aircraft. "Without a cutting-edge 'intelligent' control system, such a device would be inconceivable."
Note: For interviews, contact Professor Morse at (203) 432-4295. For more information about engineering at Yale University, see http://www.eng.yale.edu.