I wish I were born in the 1940's to sprint the space race. Humankind cannot thank enough for the legacy this bustle has left in science and technology. During my initial visit to George Mason University, I met Dr. Gerald Cook and we talked at length about his work in control systems and robotics and the type of research work I could do with him. Soon I learned that Dr. Cook had worked for the Apollo program early in his career and at that instant, I could see it clear that I wanted to work with him.
My entire career has been in the field of robotics and controls, starting from my junior year project where I developed an obstacle avoiding robot with a PLL based IR range sensor interfaced to a micro-controller. During my final year project
…show more content…
During this period I interacted heavily with the component manufacturer and various subsystems to deal with integration issues; mechanical, electrical, and thermal. I also wrote firmware to interface the spacecraft computer to the gyro, and to feed the data into the attitude estimator/controller. This work required a thorough knowledge of several fields such as the data bus, gyroscope data processing, Kalman Filters and Linear control systems. I was extremely delighted to take part in the launch operations for this spacecraft; and to see it fly in space just the way we intended it to operate, was beyond exhilarating. Three years into my career, I had the amazing opportunity to develop a complete reaction control system using control moment gyroscopes on an air-bearing platform. I had to use several skills in circuit theory, linear controls, dynamics, mechanical interfacing or computer programming for this task. Our team used a standard CMG steering algorithm and modified it to operate on the real-time testbed.
After learning complex concepts such as Multivariable and Nonlinear control, I was amazed to learn how rarely these are used in the industry. With rapid advancements computational power, I see no reason such concepts could not be applied in more and more applications. For instance, a real-time computer can be programmed to perform system identification as a housekeeping task. This, when coupled to an LQG controller can yield accurate and robust control.