Research Projects

Robot Hardware Group

The goal of our group is to develop a fleet of robots to test ideas in the fields of automotive control (intelligent cruise control), formation control, and control of large groups of heterogeneous vehicles.

We are, now, focusing on developing one robot moving autonomously, and, furthermore, to build one or two identical replicas. Developing a flexible, controllable robot, we can have a great opportunity for variety of control study. In addition, we have a plan to acquire a positioning system. We want the robots in the fleet to be able to communicate information to each other. Moreover, we are working to design a web-based interface to control and command the robots, and replay missions.

Team Control and GUI Group

We intend to outline the organization and communication formalisms for autonomous groups of robots. We consider the general case of a bottom-up approach towards robot group composition – configuring our solution to utilize present resources rather than dictate specifications for resources that need to be fabricated. Our goal is to move towards simulation and then implement our organization on our robots in the lab. We are also in charge of designing an interface for and operator that effectively and concisely displays the large amount of information associated with large robot groups. Lastly we build and customize a replay tool for our simulations and real-robot trials for post-testing analysis.


Safe Controller Design

Nowadays, unmanned vehicles are becoming part of operations in scientific and commercial applications. Safety guarantees turn to be of paramount importance during the trend.

The goal of this project is to develop safe controllers in the fields of automotive control (intelligent cruise control), formation control, and control of large groups of heterogeneous vehicles. Any control strategy could be applied for desired performance until the state reaches the boundary of the safe set, when the safe controller works for vehicle safety. The characterization of the safe sets is developed with mathematical derivative, and then numerical simulations and physical experiments on the robot test-bed are employed to validate the safe controllers.


Formation Flight Control Group

Our focus is to study control techniques, role assignment and trajectory planning that allow us to move formations of unmanned vehicles in a coordinated way. In particular, we can create and control a flight formation for unmanned air vehicles (UAV's). Applications of interest are for airplane formation flight, satellite interferometry, and detection of sources of contaminants. The order of complexity of these systems increases with the number of objects/vehicles being controlled, extension into three-dimensional space, as well as an inclusion of system disturbances such as wind. Currently we have a mathematical model in place for two-dimensional control of unmanned vehicles, which will be tested with ground robots. Another application of interest is intelligent highway systems, where vehicles can move in a specified formation and adjust accordingly given a change in velocity of the vehicle's surrounding neighbors.

 

   

Copyright Columbia University 2004