Postgraduate Study within the Robotic Systems Lab

However, there are still a wide range of fundamental questions to be studied and understood before robots can become the intelligent assistants of the future. The subject areas can be broadly divided into mobile robotics, manipulation (robot arms and hand) and perception (computer vision and understanding the environment) and potential projects in each area are presented below.

The Robotic Systems Lab is a dynamic and growing group which successfully attracts funding and is involved in a number of cooperative research projects with universities in Japan and Sweden. We endeavour to provide most PhD students with a $5000 p/a additional scholarship. In addition, students will have the opportunity to travel to at least one conference and/or to visit one of our cooperative research partners.

For information on available PhD scholarships, see the ANU Graduate School.

In addition, further information can be found on the Department of Systems Engineering Graduate Page and the RSISE Graduate Page.

Mobile Robotics

1. Robotic Operating System The use of computers really took off and started on a path of exponential growth when the operating system was developed. The operating system provides a level of abstraction from the hardware as well as basic building blocks so that developers of programs had less work to do (particularly re-inventing basic routines that others had to develop). This project seeks to design (and develop as much as possible) an operating system for robotics which provides a degree of hardware independence and basic modules which are commonly used. It is planned that this project will involve an exchange trip to Stockholm for up to four months. For more information, contact David Austin.
2. On-line programming environment The idea behind this project is to allow remote users the opportunity to develop programs for our mobile robot. At the moment, a project is underway to allow web users to use a simple point-and-click interface to command the robot to move. However, this interface is of a fairly limited nature and we are interested in developing a more fully featured interface. For example, students at high school (or college) could develop programs for the robot remotely. It is also envisaged that programs or modules developed by remote web users could be made available for other users to build on. It is planned that this project will involve an exchange trip to Stockholm for up to four months. For more information, contact David Austin.

Manipulation

1. Human-safe Robot Arm Robotic assistants of the future will have the ability to move around (see mobile robotics above) and the ability to physically interact with the environment. For example, a useful robot assistant should be able to perform tasks such as "get me a beer from the fridge", "do the dishes", and "clean up my room". However, robot arms (manipulators) that are available today are not safe for use around humans. Today's manipulators are designed for very high positioning accuracy and high speed. This means that they are heavy pieces of machinery capable of high velocities. Clearly, not something that one would want around the home. This project will focus on the design and development of a safe, lightweight robot arm which poses less of a threat to humans. It is planned that this project will involve an exchange trip to Stockholm for up to four months. For more information, contact David Austin.
2. Lightweight Robot Hand Robotic assistants of the future will have the ability to move around (see mobile robotics above) and the ability to physically interact with the environment. For example, a useful robot assistant should be able to perform tasks such as "get me a beer from the fridge", "do the dishes", and "clean up my room". However, robot arms (manipulators) that are available today are not safe for use around humans. Today's manipulators are designed for very high positioning accuracy and high speed. This means that they are heavy pieces of machinery capable of high velocities. Clearly, not something that one would want around the home. The weight of the hand or gripper at the end of the robot arm clearly has a significant effect on the safety of the arm as a whole. If the hand is heavy then the amount of kinetic energy and momentum that can be imparted in an impact is high. This project will focus on the design and development of a lightweight, multi-fingered hand. It is planned that this project will involve an exchange trip to Stockholm for up to four months. For more information, contact David Austin.
3. Mobile Manipulator Robotic assistants of the future will have the ability to move around (see mobile robotics above) and the ability to physically interact with the environment. For example, a useful robot assistant should be able to perform tasks such as "get me a beer from the fridge", "do the dishes", and "clean up my room". This project will investigate the possibility of mounting a robotic arm on our XR4000 robot, perhaps in cooperation with the projects "Human-safe Robot Arm" and "Lightweight Robot Hand". It is planned that this project will involve an exchange trip to Stockholm for up to four months. For more information, contact David Austin.
4. Modular, Reconfigurable Robot In some ways, robots are a type of meta-machine - by changing the software, we can relatively easily alter their function. This project seeks to develop a meta-robot. That is a robot which consists of a number of modules which are able to change their physical arrangement (under software control) to become a new type of robot. For example a chain of modules might act as a snake when travelling over rough terrain but might connect into a circle for rolling over smooth terrain. For more information, contact David Austin.

Perception

1. Situated Signal Processing A robot assistant of the future needs to have the ability to respond to verbal commands and queries such as "get me a beer from the fridge", "do the dishes", and "clean up my room". However, placement of a single microphone on the robot leads to unsatisfactory performance because of other sources of noise in the room. Ideally, the robot should be able to separate spoken commands from other conversations in the same room. This will require the integration of signal processing with other sensory information from the robot (such as vision, laser range scanners etc.). The vision and laser range scanners will be used to form hypotheses about where people are in the room who could be speaking to the robot and then signal processing of an array of microphones will be used to extract the speech. For more information, contact David Austin.