Spacewire - a new highspeed communications system for NASA's spacecraft, has been developed by Dr Steve Parkes and his Space System Research Group in the school of engineering.
Working with technical experts across Europe and the USA, Dr Parkes and his team have developed SpaceWire, a communications network which has become the standard for use on many spacecraft, including several developed by NASA and ESA, the European Space Agency. The onboard network connects together many of the electronic units on a spacecraft and can send information between them at high speeds. The units could be a telescope monitoring the Earth's oceans or some other type of sensor, a large memory for storing the data gathered, computers for processing the data and controlling the spacecraft, or the radio that is used to get commands and send data to Earth. Describing the system, Dr Parkes said "SpaceWire is a bit like a motorway network connecting towns together: The towns are the electronic units that have to communicate with one another and the roads are the wires connecting these units to one another." "To send information you have to put it in a packet, just like the family piling into the car."
"The difference is that instead of the driver knowing where the car is headed, a big label is put on the front of the packet stating its destination (e.g. INVERNESS)." The packet (car) is then sent along a SpaceWire link (road) until it reaches a junction called a router. The router looks at the label on the front of the packet and directs it down another SpaceWire link (road) to its destination or to another router on the way to the destination." "When the packet arrives at its destination unit the contents of the packet are extracted and delivered."
"SpaceWire is extremely reliable and has what's known as fault tolerance, which means that even when part of the network breaks down packets will still be delivered to the required destination if at all possible." "The routers help with this, when they see that a SpaceWire link has stopped working they send data along a different route to the destination, effectively putting in a diversion around the failed part of the network."
Another major facet of Dr Parkes space research is his Planet Surface Simulation project, begun in 1998 and funded by the European Space Agency. This research is real science fiction. The idea is to land an unmanned spacecraft on the surface of a distant planet like Mercury. It has to land at a precise location avoiding small craters, boulders and other obstacles that may be in the vicinity of the target landing site. The spacecraft will have a computer onboard acting as a pilot with a camera acting as the pilot's eyes. The camera will take a stream of images of the surface as the spacecraft descends and the computer will process these images to help guide the spacecraft down towards the target landing site. It will also look out for obstacles at the target landing site and adjust the descent trajectory to make sure that the spacecraft lands safely.
To test this type of system the team at Dundee have developed a sophisticated computer programme to simulate the surface of a planet and to practise the unmanned landings. This programme simulates heavily cratered landscape, similar to that found on the Moon or Mercury. The output of the surface simulator goes to a second programme, which simulates the camera looking at the surface and produces images from different positions above the surface. Another programme simulates the spacecraft and its computer pilot. Put together the simulation system can simulate the descent of the spacecraft towards the planet's surface, filming as it descends. "Obviously it's far too expensive to do these sort of tests with real spacecraft on the Moon or Mercury," Dr Parkes said, "so it makes sense, and is vastly less costly, to simulate landings using a computer." "We are constantly improving and expanding the simulation and we hope that the work we're doing will eventually lead to the first unmanned landing on Mercury."
The Courier, Dundee