14 February 2011
New research enables intelligent spacecraft to make decisions
Researchers have unveiled a control system that enables engineers to programmes satellites and spacecraft to 'think for themselves'. Scientists at the University of Southampton describe the artificially intelligent control system – 'Sysbrain' – as 'the world's first publishing system of technical knowledge for machines'. According to Professor Sandor Veres, who is leading the EPSRC funded project, the technology opens the door for engineers to publish control instructions to machines directly.
Using natural language programming (nlp) software agents can read special English language technical documents on control methods. This provides vehicles with advanced guidance, navigation and feedback capabilities and the ability to adapt during missions, identify and rectify problems and make decisions on how to best carry out a task.
To test control systems, Prof Veres and a team of engineers constructed a test facility and a fleet of satellite models controlled by the Sysbrain cognitive agent control system. A glass covered precision level table, surrounded by a metal framework was used to mount overhead visual markers, observation cameras and isolation curtains to prevent external light sources affecting experiments. Visual navigation was performed using onboard cameras, observing the overhead marker system located above the test area and replicating how spacecraft might use points in the solar system to determine orientation.
As with real satellites, the models rotated around a pivot point. To replicate the zero gravity properties of space, they were placed on the table, while gliding across on roller bearings. Each model had eight propellers to control movement, a set of inertia sensors and additional cameras to be 'spatially aware' and to 'see' each other. The model's skeletal robot frame also allowed various forms of hardware to be fitted and experimented with.
Professor Veres said: "Sysbrain is a special breed of software agents with unique features such as natural language programming to create them, human like reasoning, and most importantly they can read special English language documents in 'system English' or 'sEnglish'. Human authors of sEnglish documents can put them on the web as publications and sysbrain can read them to enhance their physical and problem solving skills. This allows engineers to write technical papers directly for Sysbrain that control the machines."