‘Eye operation’ puts Proba-1 back in business
After more than a decade in orbit, ESA’s Proba-1 was showing its age – even hibernating last winter. But a software fix to its startracker, radiation-impaired after surpassing its design lifetime fivefold, has returned the veteran Earth-observing microsatellite to full operation.
New software from the Technical University of Denmark (DTU) allows Proba-1 to distinguish between genuine star constellations to measure its pointing direction versus clusters of radiation-induced ‘hotspots’.
“The result is that Proba-1’s startracker is practically rejuvenated,” said Frederic Teston, head of ESA’s Proba programme. “The mission is back in business.”
Its name standing for ‘Project for Onboard Autonomy’, Proba-1 was designed for autonomous operations when it entered low-Earth orbit on 22 October 2001.
Ground controllers at ESA’s Redu centre in Belgium only need to supply geographical coordinates of a ground target and the microsatellite – smaller than a cubic metre – steers itself into the appropriate position, even tilting to acquire multi-angular views.
Its agility comes from its DTU-designed Advanced Stellar Camera (ASC), an autonomous CCD-based startracker whose computer takes bearings of stellar constellations to assess the satellite’s attitude and rotation.
The problem was that years of continuous bombardment by the charged particles permeating space built up bright points on the CCD that began to camouflage actual stars.
“Such hotspots correspond to single pixels, while stars extend beyond single pixels exhibiting lens effects,” explained Troelz Denver, associate professor at DTU. “So to begin with it was quite simple to distinguish the two classes.
“The challenge comes as radiation hits accumulate, having two, three or more adjacent pixels impaired. Proba-1 ended up perceiving three to four times more hotspots clusters than stars.”
The radiation effects show up more strongly on the CCD as temperatures increase. So Proba-1 went into hibernation last January, when (counter-intuitively for those in the northern hemisphere) Earth’s elliptical orbit takes it closest to the Sun – and the intensity of sunlight on the satellite increases by 10%.
But while Proba-1 was developed as a technology demonstrator with a two-year lifetime, it has evolved into an operational Earth observation mission. Hundreds of scientific teams worldwide were inconvenienced by any observation gap.
“We kept in close touch with the ESA Redu team and the mission’s prime contractor QinetiQ Space,” added Dr Denver. “A fix was possible because the ASC is completely reprogrammable in flight.
“Redu performed manoeuvres trying to decrease the startracker temperature, while they supplied imagery from the instrument for testing – we could see just what the tracker was seeing.
“The new algorithms developed made use of the fact that Proba-1’s ASC has two camera heads –preventing the Sun or Earth blocking out views of the stars along a single direction.
“When one head has difficulties isolating the stars from the hotspot clusters, it’s now being assisted by the other.”
The improved algorithm is also available for use on subsequent ASC versions, which has flown on ESA’s Smart-1 Moon mission, the Sun-monitoring Proba-2, gravity-mapping GOCE and the soon-to-be launched Swarm satellite constellation to chart Earth’s magnetic field.
DTU originally developed the ASC for Denmark’s national Ørsted satellite in the 1990s. One particular requirement was the ability to autonomously fix its orientation even when out of control –‘lost-in-space’.
Dr Denver envisages future space missions flying multiple startracker cameras: “The more heads aboard, the more complex manoeuvres can be exercised without blinding the spacecraft. And, as Proba-1 shows, the startracker’s working life is boosted.”