Embedded World: Multicore technologies enable precision space manoeuvres

Researchers from Fraunhofer Institute for Computer Architecture and Software Technology FIRST have engineered the system which can be seen at Embedded World, Nuremberg from 28 February to 1 March. For a spacecraft to 'see' and maintain its equilibrium, it needs a high performance onboard computer and the device must process a myriad of sensor data simultaneously, as well as withstanding the severe conditions of space. The MUSE project (Multicore Architecture for Sensor based Position Tracking in Space) was established to improve the positioning and guidance of such spacecraft. Under the plan, scientists from FIRST developed an extremely high performing onboard computer using modern multicore processors. High resolution cameras and infrared sensors on the spacecraft deliver immense data volumes that help determine the position of the target object. In order to compute the precise control of the vehicle, the data has to be processed in real time. Due to the high performance requirements, spaceflight enabled computers generally have to make sacrifices in terms of quality. Samuel Pletner, head of Aerospace Business Development at FIRST, said: "In space, the major challenge is this: the system must provide and enormously high computing capacity, while power supply, weight, space and cooling requirements are kept to a minimum. In addition, cosmic radiation may cause sporadic data corruption, which has to be detected and rectified by means of error tolerance mechanisms. We have to realibly eliminate the possibility of undetected errors leading to erroneous guidance commands and ultimately, uncontrolled movements of the spacecraft." The researchers claim to have resolved the problem with the P4080 multicore processor, manufactured by Freescale, which is highly integrated and robust. According to the team from Fraunhofer, the processors provide maximum processing capacity and enable more efficient error tolerance mechanisms to be achieved. The researchers have devised complex position detection algorithms especially designed for multicore architectures, so critical calculations can be conducted on a number of different processor cores and the results checked through a reliable comparison. The MUSE project is funded by the Space Flight Agency of the German Aerospace Center DLR, with funding from the German Federal Ministry of Economics and Technology. Fraunhofer will be demonstrating the technology in Hall 5, Booth 228 at Embedded World.