In the next phase of the spacecraft Rosetta's journey. the mission control team has to cope with many uncertainties--including whether the comet 67P/Churyumov-Gersaimeno might suddenly belch--like a flatulent frat boy--a plume of gasses that could induce a disturbing torque on Rosetta's attitude.
"Our biggest challenge in flying is being able to predict as accurately as possible the future orbit of the spacecraft, some three or four days out," says Andrea Accomazzo, Rosetta flight director and head of the Planetary Missions Division at the European Space Agency (ESA). He heads up the team of 28 full-time engineers responsible for the mission's flight dynamics and control.
As they prepare to deploy the lander to the Comet 67P, Andrea Accomazzo, Rosetta flight director, and his team of 28 full-time engineers, will have to contend with a devious trio of forces acting on the spacecraft and signal delays, both of which are changing continuously.
By studying images of the comet, Accomazzo says that engineers are able to reconstruct where the spacecraft was when the image was taken and use this data to predict where it will be. But he adds that there can be uncertainties of up to one kilometer in those predictions, which are made three to four days out.
And one kilometer over 20 kilometers is a significant angle of displacement, sufficient enough, he stresses, for the Philae Lander to miss the comet entirely when it attempts to land.
"Things are not so straightforward. We want to orbit this comet, so we need to know about its rotational status. But we don't know its exact gravity and we don't know its exact shape and mass, so we have to create models that allow us to make predictions on its status," explains Accomazzo.
The first phase, completed last week, was to put the spacecraft in orbit far away from the object and begin measuring the probabilities.
At a speed of mere tenths-of-a-kilometer, Rosetta is flying in front of the comet in a "hyperbolic" orbit, tracing the three arcs of a triangle. "The spacecraft is still at a velocity that is high enough not to be fully captured by the comet's gravitational field," Accomazzo explains. "However we are close enough to be affected by it so that we can detect it and measure It.. By flying in a triangle around it we can also observe the comet from different illumination angles, and we can easily construct a shape model."
What makes the control problem especially devious is the trio of forces acting on the spacecraft and the delay in signal transmissions to and from the spacecraft, both of which are changing continuously.
Three accelerations affect the spacecraft: Because of the tiny gravitational force of the comet, the aerodynamic forces created by the gasses emitted from the comet and solar radiation pressure also are in play. Currently, the gravitational force is dominating, but in a few weeks when Rosetta is closer to the comet the aerodynamic forces produced by the gasses will become fractionally more important.
While the team has relatively good models of the solar radiation pressure and have started characterizing the 67P's gravitational forces, Accomazzo admits that they do not have a very good model of the gasses emitted by it. And they might never have one, he says, because of the notorious unpredictability in the behavior of comets.
In fact, instruments have detected water vapor coming off of 67P. And though not expected, the comet, like a flatulent frat boy, might suddenly belch out a gassey plume that could induce a disturbing torque on Rosetta's attitude.
Continuously iterating on the probe's position is not an option. Though the 10-year-old Rosetta was designed in the 1990s and only has 2 MB of RAM, of which only half can be accessed by the operating system due to CPU budget constraints, technology isn't the limiting factor. Rather, it's transmission delay time and the 20-hour orbit-planning schedule, which is the time flight control engineers need to assess Rosetta's orbit, plan the future maneuvers, and finally generate the commands to implement them.
The commands are then uplinked to the spacecraft with a propagation delay (i.e., the travel time from Earth to the spacecraft) that can vary between 20 and 45 minutes. To take into account the large distances (hundreds of millions of km), the Rosetta team developed a model of the delay as the probe's distance from the earth changes, and use it to update the control system continuously.
The fact that Rosetta has traveled billions of miles to find and orbit a comet that is only 2.5 miles by 2.2 miles is extraordinary. As I understand, escape velocity from the comet is about 1 mph so if we walked on the surface, we'd "launch" off the surface, leave the comet's gravitational field, and be floating in space. It is hard to imagine the fine adjustments that must be made by Rosetta to remain in the vicinity.
@maxmaxfield :So if they'd had a larger budget, the operating system would have been allowed to access the other half of the RAM?"
I emailed Andrea Accomazzo of ESA to shed some light on your question and here is his answer:
"Yes, we could use part of the RAM extension but the CPU budget would decrease i.e. we would have the risk that the computer would not be able to run all the actions it has to run within the allocated time i.e. activities on-board would start being delayed which would be catastrophic for the mission.
Proper timing of orbital and attitude manoeuvres is essential for mission success. If the spacecraft can not execute them then we could not manoeuvre properly around the comet.
In reality the on-board computer has a protection against over-run but it would re-boot thus entering a safe status where all programmed activities would be de-scheduled i.e. the mission would be temporarily interrupted with an obvious disruption of science production and delays."
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.