My Research Activities

Research projects:

Research interests & expertise:

 

This home page is still under construction. More information will be added gradually.

Return to the top of the page

CALVIN: An Interactive Programming Language

This project is to implement a software package, Calvin, initially designed by Valiant Technology, intended to allow secondary school children to program a limited range of ATMEL processors. Calvin enables school students to tackle general control problems, but has an emphasis on robotics, particularly mobile robots. The project will also involve the development of up to 4 simple project boards that will permit students to incorporate the ATMEL processors into class projects. Some of these may link to Valiant's Tronix products, and others will be standalone designs.

Basic tasks of this project is to produce:

This project is funded by Valiant Technology Ltd. (£10,000).

Return to the top of the page

Internet-based devices and telerobots

The operation of current robots and embedded systems in industry and the service sector has remained separate and independent. In other words, these robots and embedded systems are isolated one another by different environments and have no effective way to communicate. This has made the current robots and embedded systems expensive, incompatible and long developing cycle, and seriously hampered the daily development of robot technology. Therefore, there is necessary to develop an integrated network environment for robotic systems based on today's Java computing and Internet technology.

With the rapid growth of the Internet, more and more intelligent devices or systems have been embedded into it for service, security and entertainment, including distributed computer systems, surveillance cameras, telescopes, coffee machines, manipulators and mobile robots. Although the notion of Internet robotics or web-based robotics is relatively new and still in its infancy, it has captured the huge interest of many researchers worldwide. Except for operating in hazardous environments that are traditional telerobotic areas, Internet robotics has opened up a complete new range of real-world applications, namely tele-training, tele-manufacturing, tele-surgery, museum guidance, traffic control, space exploration, disaster rescue, house cleaning, and health care.

On the other hand, Java has become the most exciting development in the later 1990's. Launched as an alternative programming language to C/C++, Java has now matured. The reduced complexity offered to developers, support for object orientation, user interfaces, and others, makes Java a good candidate for developing real-time applications in various domains of intelligent machines and robots. Java computing technology is a key for creating a network robot environment since it is a unified platform with efficiency, scalability and cross-platform compatibility. Individual applications can be designed within this environment and then executed on different computing and control devices. In other words, physical devices can be simulated "on the net", tested for correctness, then put into action. Control devices can be managed over the Internet using browser-based interfaces. In this way, process control, home and factory automation become more flexible, powerful and easier to manage. This research project is to investigate how we actually build such an environment based on Java technology, and how to implement Java-based microprocessors in the real-time control of robots and embedded systems.

The long-term goal of this project is to carry out fundamental research on multi-agent/multi-robot cooperation and learning towards real-world applications such as tele-training, tele-manufacturing, tele-repairing, remote surveillance, fire fighting, and distributed service robots for office, hospital and elderly care. Cooperative Internet robots will be a useful test-bed for us to do this. Also this unique equipment can be shared with other researchers and Internet users who are willing to work in the same area.

Further Information: URL: http://cswww.essex.ac.uk/, or contact: Prof. Huosheng Hu, Email: hhu@essex.ac.uk

Return to the top of the page

Multisensor-based Navigation and Map Building

A crucial capability for an autonomous mobile robot is the estimation of its position in the real world. It is well known that a mobile robot cannot rely solely on the dead-reckoning method to determine its position because dead-reckoning errors are cumulative. Using external sensors to observe useful features from the environment becomes necessary. There are two kinds of the environment features that can be used in localisation: one is the artificial beacon populated at a known position in a structured environment, and another is the natural beacon that is abstracted from the real world. The use of artificial beacons is to simplify the localisation process by measuring the robot motion relative to pre-placed beacons. The use of natural beacons enables mobile robots to operate in an unstructured environment, but pays a high cost to interpret sensory data and obtain useful features.

This project is to investigate how to integrate multiple simple and inexpensive sensors in the localisation process for a mobile robot to observe both features in an environment. An extended Kalman filter will be used to fuse the data from multiple sensors. As a mobile robot traverses its environment, it should be able to observe both artificial and natural beacons to update its position and the environment map continuously.

This project is funded by Robotic Sciences Ltd. (£8,000) and collaborated with NavTech Electronics Ltd.

Further Information: URL: http://cswww.essex.ac.uk/, or contact: Prof. Huosheng Hu, Email: hhu@essex.ac.uk

Return to the top of the page

Cooperation of Multiple Mobile Robots

With the fast increase in the number of robots in industry and domestic markets, scientists and engineers in robotics community are facing the challenge on the implementation of multi-robot communication and cooperation. The key factor is that a team of smaller and simpler robots can outperform a large and complex robot. The application areas of the research include a variety of tasks where either more than one robot is required to complete a task, or where a team of robots can achieve an optimal solution in terms of time, energy efficiency, and quality. Typical applications include hazardous material handling, planetary exploration, surveillance, construction, etc.

In general, a team of mobile robots to perform the specific task by cooperation has three distinguish features. First, the environment for mobile robots to confront is dynamic and multiple robots change their positions all the time. Second, multiple robots should communicate their position one another, or a central control station detects robot's position to transmit to individual robots. Third, a modular design is needed so that each component of the system can be developed systematically. This research project is to investigate how multiple robots can effectively communicate and integrate to achieve cooperation.

Further Information: URL: http://cswww.essex.ac.uk/, or contact: Prof. Huosheng Hu, Email: hhu@essex.ac.uk

Return to the top of the page