Moving a heavy vehicle by slight touching – force controlled motion system

There are many industry applications where the force between two objects needs to be precisely controlled. An application is to develop the methodologies of controlling the motion of a heavy vehicle by simply applying it with a small touching force. System analysis and simulation are needed. Based on force control theory a prototype including a forced object, a force sensor and a motion/force control system should be made and tested.

Moving target tracking for omni-directional mobile robot using stereo cameras

This project is fits in the development of methodologies of tracking a rigid target moving in 3D space (including rotational and linear motions). Some applicable examples are vision-based airplane pursuing, car chasing, mobile robot tracking, docking, and manipulations. The objective of the project is to develop a simple prototype combining of cameras, mobile robot and other device. Real-time programming, image processing, C, Linux, and basic control theory are required.

Moving target motion estimation in 3D space using stereo cameras

The project is to test and evaluate effective real-time image processing algorithms which can estimate the pose and the linear/rotational velocities of a rigid target undertaking free motion in 3D space. The knowledge of kinematics needs to be learnt. The real-time programming, Linux and C, and basic image processing are required.

Robot aircraft dynamic model identification using MATLAB

For the development of unmanned robot aircrafts, the preliminary task is to develop the effective autopilot for the autonomous flight of the aircrafts. The first step, however, is to model the motion equation (transfer functions) of the airplanes. In this project, for the real-time flight data obtained, it is to model one or two robot aircrafts using MATLAB’s System Identification and Control Tool Box. The outcome should be effective dynamic model which can be used for autopilot controller parameter tuning. You will learn basic system identification theory and techniques and become an expert MATLAB. Based on the outcome of your work, you may also need to reprogram the autopilot of our robot aircraft to witness the improved flight performance

Vision based road traffic monitoring system

This project is to further improve the performance of a prototype of a vision based traffic monitoring system developed in our lab. The system developed, using the low cost web camera and standalone PC, has the basic functions of detecting the presence of vehicles on road, counting their number, detecting their speed, and real-time traffic data collection. However, the reliability of the system drops when traffic flow increases and the lighting condition varies. This project is aimed at enhancing the robustness of the system by means of upgrading the current machine vision algorithms, introducing more advanced functions and improving its tracking capability. The skill of C programming is needed in Windows environment.

Developing micro-gravity environment for space robots

For efficient training of space robot manipulation on the earth, it is substantial to make a virtual zero-gravity or micro-gravity environment. The major challenge of course is to cancel the gravity force applied to both robots and working pieces to be manipulated. This project is to investigate the feasibility of applying vertical-tension-suspension system to make this virtual environment. The system compensates the gravity force by applying a tension in the opposite direction of gravity force applied to the rigid body to be considered. The control system should ensure the static and dynamic compensation of the gravity force. The study should be done by undertaking simulation and then practical justification if possible.

Vision/laser data fusion for traffic monitoring vehicle

A vehicle with on-board sensor including laser range finder and stereo camera used for monitoring the motion of other vehicles around is to be established. The project is to analysis the data and signals from vision and laser to estimate the relative motion between the vehicle and the traffic around. The system can also be used for the study and modeling of drivers’ driving behavior. Some software modules have been developed and major task is to combine the vision and laser signal together in real-time to estimate the relative speed and position of the motion objects. You must like to face the challenges of C programming and basic image processing.

IMU based motion detect system

Following the advancing of SMS technique, small Inertial Measurement Unit’s (IMU) that provide accurate monitoring of linear acceleration and angular rate have become available with reasonable price. This kind of units can be easily embedded onto the moving objects for their motion detection with relatively lower cost. The examples including the motion measurement of the robot arm, human body, airplane, mobile robots, and transport vehicles. The detected motion signal can be easily transferred to a central unit either through wired-link or wireless connection, such as blue tooth or wireless internet, for information acquisition and data fusion. In this project, we are to build a very basic system applying the IMU for the motion detection for some typical devices such as a robot arm, a mobile robot or a human body. The challenging of the project is to implement effective filtering due to the exciting of the noise signal