Your questions answered: Rosetta and Philae

12 min read

Comet calls: Our panel answers questions on the recent Rosetta mission, which landed a probe on a 4km-wide comet

Landing a spacecraft the size of a washing machine on a 4km-wide comet billions of miles away is one of the most notable feats in the history of  space exploration. Questions from readers for our team of experts from the Rosetta mission flooded in; we’d especially like to thank the students of the STEM Academy Rocket Club and Year 13 Physics Group of the Buttershaw Business and Enterprise College in Bradford. Thanks are also due, of course, to the panel.

  • Phil McGoldrick, engineering manager for Rosetta, Airbus Space & Defence, Stevenage
  • Steve Kemble, senior mission analyst, Airbus Space and Defence
  • Dan Andrews, co-investigator, Ptolemy mass spectrometry instrument on Philae, Open University
  • Nick James, BAE Systems engineer, leader of Intermediate Frequency Modem System (IFMS) team
  • Laurence O’Rourke, ESA science operations coordinator and lander systems engineer


We heard the sounds generated by the lander. Given that sound doesn’t travel in a vacuum, exactly how were these sounds generated, what do they tell us and why is sound the best way to understand that data?

PM: The sounds are picked up by the feet of the lander. The sensors are in the feet — an experiment called SESAME (Surface Electric Sounding and Acoustic Monitoring Experiment). It picks up the vibrations at acoustic frequencies.

DA: While sound and vibration doesn’t travel through the vacuum of space, it is still free to travel through other media such as structures. This is how it’s possible to hear, for instance, thruster firings or docking mechanisms engaging from inside the ISS. What was recorded at touchdown were the vibrations as the feet first pushed through a layer of softer material before striking a hard surface — and the best way for humans to sense such high-frequency vibrations are as sound.

Comet 67P Churyumov-Gerasimenko, imaged from Rosetta

It took 10 years to get Rosetta to the comet, which means all the on-board systems are more than a decade old.

a How much more could you achieve if you had up-to-date tech out there?

NJ: The IFMS on the ground is a software-defined radio and so can be adapted to handle many different formats. Unlike modern commercial software one of our key requirements is backwards compatibility with old systems that are already flying.

PM: More up-to-date tech could be chips that could store more data — and therefore more data could be sent back by Rosetta in each transmission. Increased memory would also allow more frequent measurements to be taken — before the chip became full, and had to transmit data.

DA: Designing a clean-sheet spacecraft today we could use more advanced cameras and other instrumentation, new materials, new propulsion techniques, new communications techniques and probably do the whole thing even more efficiently. Yes, we could potentially get better data. I say potentially though, because this is only part of the story. Reliability is also crucially important in any space mission and particularly one such as Rosetta with the long timescales involved. Thus the use of ‘heritage’ hardware with proven reliability is one of the sacrifices that is well worth making to gain confidence that it will survive the rigours of long-duration space exploration. There is another side effect of the rapid technical progress we are seeing; in 2030–40 when a hypothetical ‘Rosetta 2’ arrived at its target, we’d still be asking the same questions: comparing then outdated 2010s technology to the technology people are carrying around in their pockets.

Layout of instruments on Philae

b: How frustrating is it to be working with technology you know is inferior to the stuff you could send now?

PM: It is not frustrating as all space technology needs to be robust, for example, to resist charged particles and the problems of bit flips — so for us it is business as usual. There is no point in sending the latest systems if they are vulnerable to failure

DA: To be perfectly honest, it isn’t. The wealth of data coming back and the success of the mission to date is beyond my wildest dreams. Also, with the Ptolemy instrument, we had the foresight to design an instrument that could be reprogrammed at the comet, to take advantage of everything we have learned since launch and such modifications were actually made at short notice in response to the ‘non-nominal’ landing


c: Are there any compatibility issues between the 10-year-old tech in the sky and the presumably more modern tech on the ground, or is mission control still using Windows Vista?
PM: Again it’s business as usual. For example, our sophisticated telecommunications satellites are designed for a minimal lifetime of 15 years in space — so we are experienced at running spacecraft successfully over long periods. One of our satellites has so far clocked up more than 22 years of successful in orbit operations.

DA: None whatsoever. A greater challenge is in people management — keeping teams together in the current funding climate, either keeping individuals or at least their knowledge base current throughout the duration of the mission. This is where good data management practices and archives such as the Planetary Data System (PDS) come in.


How was this particular comet selected?

DA: Comet 67P/Churyumov-Gerasimenko wasn’t actually the initial target comet. Rosetta was originally intended to launch in 2003 and travel to comet 46P/Wirtanen (another comet in the Jupiter group) but sadly the Ariane 5 launch immediately prior to what would have been Rosetta’s turn failed and it was decided not to risk Rosetta on the same type of booster until the cause of the failure was known. This led to a one-year-delay and a new target, chosen both due to scientific interest and also that we could get to it with the delta-v available without incurring too much of a launch delay.

LOR: There were both engineering (flight dynamics: capability to be able to reach a target from an Ariane 5 launch and within a certain number of years) and scientific (knowledge of the properties e.g. size and activity of the comet) factors taken into account for its selection.


How complicated are the calculations for plotting Rosetta’s intercept route and what was the margin of error?

SK: These are the result of detailed mathematical simulations of the orbital dynamics and the manoeuvres, combined with trajectory optimisation techniques to minimise fuel. The comet is approximately 4km in diameter and the distance travelled over a billion kilometres, so the percentage error in the calculations and their implementation via navigation techniques is very small.

Rosetta's orbit around Comet 67P

How long can Rosetta remain functional in orbit around the comet and what data are you hoping it can send back, even if Philae remains dormant?

SK: The orbiter is planned to remain active until around the perihelion passage (i.e. approximately 10 months, and then also slightly past perihelion. Flying in close proximity to the comet, the orbiter possess a suite of instruments to observe and analyse the comet’s properties. Also, the Rosetta mission was specified to operate until December 2015 — and it therefore has plenty of fuel on board.

DA: By December 2015, Rosetta will have observed comet 67P’s activity increasing as it approaches the Sun and reacting in ways that we simply cannot yet predict with any certainty. That is by no means a hard-and-fast end. As is typical with exploration missions, Rosetta may live on as it again recedes from the Sun, its available power reducing as sunlight grows dimmer and activity likely reducing to reflect its quarry

LOR: Rosetta can remain functionally around the comet until August/September 2016. This time period is derived from two main operational aspects; the first is the fuel that is expected to be depleted by then, the second is the distance from the sun. At that time, the distance becomes comparable to that where Rosetta had to go into hibernation a number of years ago.


Realistically, what are the chances of being able to revive Philae?
LOR: There are two main drivers for a Philae re-awakening. The first is the power; here I believe that Philae has a good chance to revive itself if you recognise that the larger solar panel has been oriented to the Sun (before it went to sleep) and that we are getting closer to the Sun so the power it generates will increase over time. The second is the thermal environment. The lander unit/science operations can only start above -40°C, thus when Philae wakes up the first thing it has to be able to do is switch on its heaters to heat up the units (and the batteries) to operating temperatures. Philae can only revive itself if its temperature environment does not get too cold e.g. below -55°C. As Philae was effectively off and did have quite low temperatures during the hibernation phase, there is a high possibility that even well below these low temperatures it should still work.

PM: Very good and we are hopeful.  When Philae landed, the comet was three times further away from the Sun than Earth. By March next year it will be same distance as Earth and will have 10 times more solar energy. So even in the shadows there will be more light, and the Sun angle will change, which should also help. Like most spacecraft, Philae went to sleep when power levels dropped. And like most spacecraft it is designed to come back to life when power levels rise — all power will be sent to the batteries until they have enough charge to wake up the craft again.

Philae is prepared for launch in 2004

• As the comet perihelion is passed and the comet heads off back for the outer reaches of the solar system, how long can mission control remain in contact with Rosetta and (if it’s revived) Philae? Could the mission last as long as Voyager?

SK: In principle an extended mission phase after perihelion is possible. The comet is likely to ‘outgas’ near perihelion, which could potentially result in damage to the orbiter, but the orbiter will ‘stand off’ to minimise damage.

DA: The final end to the mission will ultimately be governed by dwindling levels of sunlight and fuel. In the same way that Rosetta had to hibernate during its cruise to the comet due to lack of available power, the same power reduction will come into play as Rosetta recedes from the Sun again. While the active mission won’t last as long as Voyager, Rosetta itself will eventually become ‘just another tail particle’.

LOR: To go into hibernation and come out again requires fuel. In that respect, it won’t last as long as Voyager because the fuel is the driving factor for the end of the Rosetta mission.

 ’In principle an extended mission phase after perihelion is possible.

Steve Kemble, Airbus Space & Defence

How much does Philae actually ‘weigh’ while on the surface of the comet ?

DA: The comet has a mass of 1.0 +/- 0.1 x 10^13 kg, or roughly 10 billion tons — about the same as a small mountain. The size and mass of the comet yields a tiny acceleration due to the fact that gravity at the surface is some 50,000 times lower than on the surface of Earth. Don’t forget though, Philae still had 100kg worth of momentum to be absorbed by the landing gear.


Did the lander’s software alter anything about the approach, landing or how the mechanism of landing works because of the failed thruster?

PM: No. The feet have a spring dampening system for the landing.

NJ: The lander had no control over its trajectory. This was determined entirely by the ejection from Rosetta. It is likely that the failure of the ADS (active descent system) thruster contributed to the first bounce.


What parts of the lander’s experiments were completed fully, what parts only partially (and what was different in comparison to the plan) and what parts didn’t execute at all?

LOR: All instruments were executed on the comet. In certain cases, a sub-instrument was not used e.g. CIVA microscope, or results were inconclusive e.g. SD2, APXS.

DA: Ptolemy made a serious of atmospheric ‘sniff’ measurements starting just nine minutes after the initial bounce and was later able to analyse one sample oven containing a molecular sieve designed to trap down coma material. Ptolemy did not receive a solid sample — it was only possible to drill one sample during the modified First Science Sequence and this was delivered to the COSAC instrument.


Could landing the probe in any way alter the course of the comet and if so is this calculable?

SK: The mass of the lander is relatively small compared to the comet mass and the impact speed very small, so the momentum transfer is negligible and the change in the comet orbit is undetectable.


How did you plan the trajectory for Rosetta? Since it is an object-tracking problem, did you use a feedback controller or was it all planned in advance and done via feed-forward commands?

SK: In the final approach phases the comet is acquired visually by the orbiter and relative position and velocity estimates are improved. A sequence of manoeuvres is then used to progressively approach the comet, with further measurements taking place between manoeuvres. The process takes several weeks so it is not an automatic control loop, more of a ‘man-in-the-loop’ system with ground planning to execute the control.

NJ: Rosetta is moving under the gravitational influence of the Sun, the comet and the other bodies in the Solar System. Its trajectory is also affected by non-gravitational forces (thermal radiation from the spacecraft, thruster firings, drag from the comet’s coma and so on). The IFMS on Earth allows very precise measurement of radial velocity and range and position on the sky. These measurements, along with the gravitational and non-gravitational force models, are used by the Flight Dynamics team at ESOC to compute the precise trajectory. Since the mass of the spacecraft and the thruster efficiency is known it can determine the length and direction of thruster burns required to adjust the course. The ‘feedback’ then consists of doing further measurements and modelling of the orbit

LOR: From an operational perspective, the trajectory of Rosetta was already planned before launch in 2004 based upon the new comet selected. The only mode where Rosetta does use a feedback controller is in the asteroid-tracking mode where the navigation  camera feeds in what it sees into the control loop. Every other mode is through commanding from ground.

The Rosetta mission lifts off from Kourou aboard an Ariane rocket

What are the challenges involved in receiving such a weak and very narrow beam signal from Rosetta and especially when the signal is arriving at the Earth after 45 minutes of travel in space? Are highly directional ground station antennas continuously tracking/moving in Azimuth and Elevation to search for such a signal?

NJ: The antenna is pointed very precisely using predictions of the spacecraft’s position on the sky (this is called autotrack). The IFMS can also provide feedback to the antenna pointing controller to allow it to maximise the available signal level. The signal is very weak by the time it is received by the IFMS so we use very narrow bandwidths (a few tens of Hz on the carrier and milli-Hz on the ranging signal) to extract the synchronisation information. This is assisted by an FFT that estimates the carrier frequency.

LOR: The challenge primarily comes from the size of the ground station antenna used on Earth to acquire the Rosetta signal. Normally you need 35m or greater at the distance where Rosetta
is. There is no need to do any search for the signal as Rosetta’s position is very well known and calculated. Currently the one-way travel time of the signal is about 28 minutes. For low-Earth-orbit satellites the ground station antennas do indeed move to maintain contact with them.

The Malarguee tracking station, which is receiving Rosetta's signals

What is the thermal-control system that is used in Rosetta and Philae? Is there any difference between the system used in both
of them?

PM: Rosetta has radiators to get rid of excess heat and heaters to ensure the spacecraft does not get too cold. Rosetta also has a series of louvres that open and close automatically to maintain the correct temperature.

LOR: Rosetta has an active thermal-control system that maintains temperatures of all units within a certain predefined set of limits. If the temperature gets too low, it switches on a heater, if it is too high then it switches off a heater and takes advantage also of a passive thermal control to aid in cooling. For Philae, it has two passive absorber foils mounted to the lid, collecting solar energy and dissipating the transformed thermal power via radiation into the internal compartment  It also has electrical heaters. Contrary to Rosetta itself, Philae has the capability of maintaining the temperature of its units ‘above’ a certain temperature but it has no way to reduce a temperature besides switching off units. The three-month lifetime on the comet (at least for the original landing site) was linked to the temperature within the lander getting too high at the end for the lander to be able to sustain operations.


What attitude-control systems does Rosetta have and what are their respective roles?

PM: Rosetta has a bi-propellant thruster system, which is the same technology that we use on our telecommunications satellites, so flight proven and reliable.

DA: Rosetta has two attitude-control systems, one using flywheels and one using thrusters. Most attitude control is done via flywheels which are spun; any increase or decrease in the rotation speed of a flywheel will cause an opposing motion in Rosetta via the conservation of angular momentum (in the plane of the flywheel). Since flywheels act only in one plane, several are needed to control the spacecraft’s rotation in 3D space. These flywheels do however have limits and need to be ‘offloaded’ if they start to approach operating limits of RPM — these offloads are accomplished by firing bipropellant mono-methyl hydrazin thrusters in such a way as to apply a torque to the spacecraft, countered by spinning down the relevant flywheels. All delta-v manoeuvres are all carried out by the thruster system, a series of 24 10-Newton thrusters, four of which (the ‘axial’ thrusters) point along the Z-axis of the spacecraft for this purpose.

LOR: Rosetta’s thruster system is normally not in use as fuel is a life-limited item on the spacecraft. In that respect, it is used when the spacecraft is in safe mode or indeed when it performs manouevres. In all other modes, it uses its three reaction wheels linked to the degrees of freedom of the satellite and controls its attitude using these. When the reaction wheels reach a certain velocity, then the thrusters are used to offload the momentum that has built up in the wheels.


What systems are active in Rosetta’s hibernation mode?

DA: Thermal systems were active to make sure the spacecraft did not freeze and the receiver was active, waiting for the signal to tell it to wake up.

LOR: Primarily the main computer and the thermal-control system.

 ’Launching RTGs involves launching plutonium and that would be a problem for Europe

Nick James, BAE Systems

Why did you choose solar panels over, say, an RTG?

DA: Solar panels are simple and reliable, and proven on our previous scientific and commercial satellites. They will generate a lot of power as Rosetta travels nearer to the Sun.

NJ: Mainly political in that launching RTGs involves launching plutonium and that would be a problem for Europe. ESA is investigating RTGs that use a different isotope and this may allow it to launch in future. US spacecraft also use small amounts of radioactive isotopes to provide heating and this would have been useful for Philae.

LOR: ESA does not use any nuclear capabilities for its spacecraft. In that respect, we were forced to build large solar panels to give the required power.