by Janice Hamilton

When the next generation of Canadian satellites and interplanetary space probes are launched, technology developed at Concordia could be on board.

A multi-university, multidisciplinary team headed by Professor Khashayar Khorasani, of the Department of Electrical and Computer Engineering, is developing technology that will allow spacecraft to diagnose and recover from on-board problems by themselves. The project is funded by an $830,000 strategic grant from NSERC and $104,000 from the Canadian Space Agency.

“This is an ambitious project, but it is feasible,” Khorasani said. “We have a good team, and the members complement each other well.”

Khorasani’s expertise is in control and intelligent systems. The other Concordia team members, all of them with the Depart- ment of Electrical and Computer Engineering, are: Assistant Prof-essor Purnendu Sinha, an expert in fault-tolerant computing; Assistant Professor Shahin Hash-trudi Zad, an expert in fault diagnostics; and Associate Professor Ferhhat Khendek, an expert in real time systems. The other members of the team are from Simon Fraser University, the University of Western Ontario, and the University of British Columbia.

“This project is quite multidisciplinary in nature, and covers a wide variety of issues,” Khorasani said. “The challenge is to put these areas of expertise together in order to look at the problem of diagnosis and intelligent control in a holistic manner.”

One of the difficulties facing space vehicles, especially those on interplanetary missions, is that they are so far away from Earth that radio transmissions to and from ground-based controllers can take hours. This project, titled Intelligent Autonomous Space Vehicles (IASV): Health Monitoring, Fault Diagnosis and Recovery, will lead to the development of space vehicles that don’t have to wait for instructions from Earth if something goes wrong.

“The idea is to develop on-board autonomy for the system. The satellite does all the diagnostics, or identification of problems, on its own, and comes up with solutions to rectify the problems. But you are dealing with an uncertain environment, and there are many things you cannot plan in advance, so you have to use artificial intelligence systems and techniques to be able to plan and schedule remedies if there is a problem.”

Space vehicles have low-level sub-systems and components, such as sensors, actuators and propulsion systems, that do individual tasks, as well as high-level systems responsible for planning and scheduling.

Khorasani gives the analogy of an institution in which low-level employees carry out tasks and managers supervise. When a problem with one task causes problems in other areas, managers may have to re-allocate tasks to maintain performance. Similarly, high-level controllers in autonomous space vehicles have to supervise the interaction of low-level components.

In this project, Sinha, Khendek and Hashtrudi Zad will deal with high-level issues, while the other team members concentrate on low-level issues. Then everything has to be put together as a complete and integrated system.

The Concordia team became involved in this project after working on a smaller research contract with the Canadian Space Agency.

The four-year project, which got underway in October, will involve the training of 20 students at the masters, PhD and post-doctoral levels. These students, now being recruited, will have the opportunity to work with people at the Canadian Space Agency and see how missions are conducted.

Khorasani says this technology may be used in Radarsat-2, a Canadian imaging satellite to be launched in 2004 for mapping, environmental monitoring and other applications. For example, if this satellite detects major fires on Earth, it could rapidly change its orbit to monitor them, thereby increasing its scientific usefulness.

Aircraft engine manufacturers are also interested in using these diagnostic tools in their engines to do in-flight data diagnosis.