Paolo Ferri, Head of Mission Operations at ESOC Paolo Ferri has been a colleague and friend over a number of years now, since he was a contributor to the Space Systems Engineering courses I used to organise (prior to my retirement in 2010) at ESA Estec, ESA’s Technical HQ in the Netherlands. As Head of the Mission Operations Department at ESA’s European Space Operations Centre (ESOC) located in Darmstadt, Germany, he has managed the team involved in the Rosetta mission since its launch in 2004. Consequently, he has fronted many of the recent PR events as a consequence of the amazing success of the Rosetta spacecraft and Philae landing missions. News of these events has reverberated around the world, and made a huge PR impact for ESA, which is very much well-deserved – Rosetta in my view is the most daring and complex robotic mission ever.
I always think of Paolo as a good friend, but also a very humble guy. Now that he has established his place in the astronautics history books, I’m sure he will handle his new-found fame with good humour and humility as he has always done in the past. Characteristically, despite the fact that he must be very tired after his recent very long working days (did he get any sleep in the couple of days leading up to the Philae landing on 12 November?), he has agreed to provide an input to this blog, for which I am very grateful. The format is along the lines of questions (me) and answers (Paolo), so I hope you enjoy this brief insight into recent events from the ‘horses mouth’, so to speak!
I always think of Paolo as a good friend, but also a very humble guy. Now that he has established his place in the astronautics history books, I’m sure he will handle his new-found fame with good humour and humility as he has always done in the past. Characteristically, despite the fact that he must be very tired after his recent very long working days (did he get any sleep in the couple of days leading up to the Philae landing on 12 November?), he has agreed to provide an input to this blog, for which I am very grateful. The format is along the lines of questions (me) and answers (Paolo), so I hope you enjoy this brief insight into recent events from the ‘horses mouth’, so to speak!
GS: Looking at the mission overall, what were the most critical events?
PF: The two things that really worried me beforehand were the hibernation phase and the initial comet phase, the former since we had to leave the spacecraft alone for 31 months, without any contact. No mission had done this before its primary scientific phase. Also, Rosetta is a 3-axes stabilised spacecraft, and to leave it alone we had to spin it up so that we could deactivate the on-board attitude control. Finally we were at distances from the Sun that no solar-powered spacecraft had ever reached before. The reactivation on 20th January this year was probably the most crucial moment in the whole mission – no signal would have meant a complete loss of the mission. Receiving the signal was big relief and a great joy!
The initial comet phases were also critical because once again Rosetta had to do something no one had tried before – flying in the proximity of a comet. This is very different from normal spaceflight. The comet is a small object, with a faint and very irregular gravity field. Additionally it also has gas and dust around it that escapes with high velocity. We had to create a model of the dynamical environment and of the comet (mass, gravity potential, rotation axis, centre of mass, gas and dust flow dynamics, surface characteristics, etc.) such that we could properly fly around it. And to build this model we had to fly around it!
GS: The rendezvous manoeuvre with the comet was the first such event for the Rosetta mission, after the previous fly-bys. What were the special challenges?
PF: The rendezvous manoeuvre after the exit from hibernation was large (a ‘delta-V of about 800 m/s was required) and had to be done with small thrusters (4 thrusters, each with a thrust of 10 Newtons). So we had to split it into many burns (during the period May to July 2014), the longest of which was more than 7 hours. Before hibernation we had done a similar manoeuvre and had experienced several anomalies in the behaviour of the thrusters, so we were very worried this time. This concern was made more acute since now the thrusters were performing in worse conditions (the feed pressure in the propulsion system now being lower than before). However, thanks to a re-tuning of the thruster controller software that we performed in February, everything went very smooth, much to everyone’s relief.
PF: The two things that really worried me beforehand were the hibernation phase and the initial comet phase, the former since we had to leave the spacecraft alone for 31 months, without any contact. No mission had done this before its primary scientific phase. Also, Rosetta is a 3-axes stabilised spacecraft, and to leave it alone we had to spin it up so that we could deactivate the on-board attitude control. Finally we were at distances from the Sun that no solar-powered spacecraft had ever reached before. The reactivation on 20th January this year was probably the most crucial moment in the whole mission – no signal would have meant a complete loss of the mission. Receiving the signal was big relief and a great joy!
The initial comet phases were also critical because once again Rosetta had to do something no one had tried before – flying in the proximity of a comet. This is very different from normal spaceflight. The comet is a small object, with a faint and very irregular gravity field. Additionally it also has gas and dust around it that escapes with high velocity. We had to create a model of the dynamical environment and of the comet (mass, gravity potential, rotation axis, centre of mass, gas and dust flow dynamics, surface characteristics, etc.) such that we could properly fly around it. And to build this model we had to fly around it!
GS: The rendezvous manoeuvre with the comet was the first such event for the Rosetta mission, after the previous fly-bys. What were the special challenges?
PF: The rendezvous manoeuvre after the exit from hibernation was large (a ‘delta-V of about 800 m/s was required) and had to be done with small thrusters (4 thrusters, each with a thrust of 10 Newtons). So we had to split it into many burns (during the period May to July 2014), the longest of which was more than 7 hours. Before hibernation we had done a similar manoeuvre and had experienced several anomalies in the behaviour of the thrusters, so we were very worried this time. This concern was made more acute since now the thrusters were performing in worse conditions (the feed pressure in the propulsion system now being lower than before). However, thanks to a re-tuning of the thruster controller software that we performed in February, everything went very smooth, much to everyone’s relief.
Joy and relief in the Main Control Room at ESOC as Philae's touchdown is confirmed - Paolo is on the right hand side of the picture. GS: Is the spacecraft in good shape after its 10 year journey?
PF: The spacecraft is very healthy. We had two major problems in the early years:
1) a leak in the high pressure part of the propulsion system. This was isolated, but it prevented us from attempting a re-pressurisation of the tanks for the recent rendezvous manoeuvre. This meant that we have to carry out the rest of the mission in “blow-down mode” – that is, the operating pressure of the thrusters decreased with the fuel consumption (as the tanks contain less fuel and the gas that provides the pressure expands to a larger volume). For the time being the thrusters are working very well at around 7 bar of pressure (they were designed to work at 17 bar and are qualified down to 10 bar).
2) two of the reaction wheels started to show a noisy behaviour that is known as "cage instability". Over the years of hibernation, using an engineering model on the ground, we developed a new strategy for the operation of the wheels at low speeds, and this seems to be working, with no sign of further degradation over the past year. Also, just in case, we have developed a back-up attitude control mode that uses 2 wheels and thrusters, in case two of the four available wheels break down.
GS: What did you think when you first saw the comet nucleus close up?
PF: I can't describe the feelings. My first rational impression, though, was that this surface is ‘alive’, and very young – that it keeps changing shape and appearance very quickly. This is plausible, as the processes that change it are linked to the revolution period around the Sun, which is only 6.5 years. I am convinced the surface will change rapidly in the coming months and we will see a different comet after perihelion (closest approach to the Sun) in August 2015 compared to the one we see today.
GS: The comet orbit phase is unusual due to the very 'low energy' spacecraft dynamics. Again what were the special challenges?
PF: See my explanation above related to the comet modelling. Another challenge was the precise navigation required. Normal radio frequency measurements are not enough to navigate around the comet, since what counts is the relative position and velocity between Rosetta and the comet, and the comet has no radio transponder! So our flight dynamics experts had to develop optical navigation techniques. They used images taken by the on-board cameras to identify and recognise landmarks on the surface. They continuously compared the landmark positions over successive pictures and managed so to perform "triangulations" that allowed us to reach the necessary navigation accuracy.
PF: The spacecraft is very healthy. We had two major problems in the early years:
1) a leak in the high pressure part of the propulsion system. This was isolated, but it prevented us from attempting a re-pressurisation of the tanks for the recent rendezvous manoeuvre. This meant that we have to carry out the rest of the mission in “blow-down mode” – that is, the operating pressure of the thrusters decreased with the fuel consumption (as the tanks contain less fuel and the gas that provides the pressure expands to a larger volume). For the time being the thrusters are working very well at around 7 bar of pressure (they were designed to work at 17 bar and are qualified down to 10 bar).
2) two of the reaction wheels started to show a noisy behaviour that is known as "cage instability". Over the years of hibernation, using an engineering model on the ground, we developed a new strategy for the operation of the wheels at low speeds, and this seems to be working, with no sign of further degradation over the past year. Also, just in case, we have developed a back-up attitude control mode that uses 2 wheels and thrusters, in case two of the four available wheels break down.
GS: What did you think when you first saw the comet nucleus close up?
PF: I can't describe the feelings. My first rational impression, though, was that this surface is ‘alive’, and very young – that it keeps changing shape and appearance very quickly. This is plausible, as the processes that change it are linked to the revolution period around the Sun, which is only 6.5 years. I am convinced the surface will change rapidly in the coming months and we will see a different comet after perihelion (closest approach to the Sun) in August 2015 compared to the one we see today.
GS: The comet orbit phase is unusual due to the very 'low energy' spacecraft dynamics. Again what were the special challenges?
PF: See my explanation above related to the comet modelling. Another challenge was the precise navigation required. Normal radio frequency measurements are not enough to navigate around the comet, since what counts is the relative position and velocity between Rosetta and the comet, and the comet has no radio transponder! So our flight dynamics experts had to develop optical navigation techniques. They used images taken by the on-board cameras to identify and recognise landmarks on the surface. They continuously compared the landmark positions over successive pictures and managed so to perform "triangulations" that allowed us to reach the necessary navigation accuracy.
The Philae lander, imaged from the Rosetta spacecraft after separation. GS: Why did you choose such a long descent duration for Philae, given that the unpredictable disturbances caused by the comet’s environment build up more over the longer time period?
PF: There were two main reasons:
1) the lander separation mechanism is designed to impart a variable ‘delta-V’ to the lander in the range 0 to 100 cm/s (well, more or less in this range). However it also has an emergency mechanism, based on springs, that was designed to ‘jettison’ the lander in case the prime mechanism did not work. This emergency system gives the lander a fixed ‘delta-V’ of 18 cm/s. Following intense discussions in the last year the lander team requested us to find (if possible) landing strategies that would require a delta-V of 18 cm/s, such that, in case the prime mechanism did not work, the emergency system would still provide the lander with the same ‘correct’ ‘delta-V’ for the landing. Given this decision, we had to find a trajectory that allowed landing with a small (18 cm/s) difference in velocity between the lander and the orbiter at separation. At the same time we did not want Rosetta to perform the separation on a ‘kamikaze’ orbit – i.e. one that would bring the spacecraft to a collision with the comet in case something went wrong with the planned post-separation orbit manoeuvre. So we had to separate very far away from the surface (22.5 km), such that Rosetta flew no closer than 2.5 km from the surface in case of post-separation problems.
2) the second reason is related to the accuracy of our navigation. You are correct in saying that a longer descent will accumulate larger errors in the lander trajectory. However the errors in the Rosetta navigation, i.e. the accuracy with which we can predict the position and velocity of Rosetta at the time of separation, are much smaller if we fly at larger distances from the comet. So, in the end the higher altitude reduces the size of the landing error ellipse (the largest contributor to it being the accuracy of the position, velocity and direction of the lander at separation time, which are all dictated by the Rosetta orbit).
So, that's the reason why in the end we had to drop the lander from such and incredibly large distance. It wasn’t easy, but our Flight Dyamics people are incredible - they assured me that we wouldn’t miss the comet, and I trusted them!
PF: There were two main reasons:
1) the lander separation mechanism is designed to impart a variable ‘delta-V’ to the lander in the range 0 to 100 cm/s (well, more or less in this range). However it also has an emergency mechanism, based on springs, that was designed to ‘jettison’ the lander in case the prime mechanism did not work. This emergency system gives the lander a fixed ‘delta-V’ of 18 cm/s. Following intense discussions in the last year the lander team requested us to find (if possible) landing strategies that would require a delta-V of 18 cm/s, such that, in case the prime mechanism did not work, the emergency system would still provide the lander with the same ‘correct’ ‘delta-V’ for the landing. Given this decision, we had to find a trajectory that allowed landing with a small (18 cm/s) difference in velocity between the lander and the orbiter at separation. At the same time we did not want Rosetta to perform the separation on a ‘kamikaze’ orbit – i.e. one that would bring the spacecraft to a collision with the comet in case something went wrong with the planned post-separation orbit manoeuvre. So we had to separate very far away from the surface (22.5 km), such that Rosetta flew no closer than 2.5 km from the surface in case of post-separation problems.
2) the second reason is related to the accuracy of our navigation. You are correct in saying that a longer descent will accumulate larger errors in the lander trajectory. However the errors in the Rosetta navigation, i.e. the accuracy with which we can predict the position and velocity of Rosetta at the time of separation, are much smaller if we fly at larger distances from the comet. So, in the end the higher altitude reduces the size of the landing error ellipse (the largest contributor to it being the accuracy of the position, velocity and direction of the lander at separation time, which are all dictated by the Rosetta orbit).
So, that's the reason why in the end we had to drop the lander from such and incredibly large distance. It wasn’t easy, but our Flight Dyamics people are incredible - they assured me that we wouldn’t miss the comet, and I trusted them!
First touch down site from a range of 40 metres GS: What are the most significant future events from an operations point of view?
PF: After the end of the Philae relay phase, Rosetta will stay in a 20 km orbit for several weeks. Then, depending on the comet activity, it will increase the distance but continue following the comet in its journey towards the Sun. After perihelion in August 2015, we will try to go closer again as the activity decreases.
GS: What will we learn from the science?
PF: Well, it's not to me, the bus driver, to say what my passengers are up to! However in general Rosetta is about learning everything about comets. And since these objects are among the oldest objects in the solar system, the aim is to attempt to understand from what type of material the solar system was made, how it evolved, and how the planets formed. Also, are comets the source of the abundant water on the surface of the Earth, and maybe of life?
GS: ESA has truely 'come of age' with missions like the Titan landing in 2005, and now the Rosetta mission, which I rate as one of the most complex robotic missions. Why, in your view, does ESA not get the same kudos as NASA in the general sphere of astronautics? Is it a deficit in the ESA PR effort?
PF: I think Rosetta has changed this situation in the past few days. The resonance around the world, even in the US, was incredible.
GS: What does current analysis tell us about what actually happened at touch-down?
PF: We are learning more and more about this. We also have incredible pictures that show the lander flying over the surface after having touched down once! We are also close to finding the exact place Philae ended up after its two long jumps (it is very far from the initial touch down point!). What happened is simply that the harpoons did not fire at all. So Philae bounced, keeping its attitude since the flywheel was still rotating. However the first touch down on the surface automatically stopped powering the flywheel, which slowly reduced its rotation, thereby transferring angular momentum to the Lander body, which started rotating. After the first long jump that lasted almost 2 hours, Philae touched down once more and bounced again, but this time for a very short jump of 6-7 minutes, before stopping in its final position, most likely leaning against a rock or the wall of a crater. Fortunately we never lost the signal, and we had all the 5 foreseen contacts during the 2.5 days on the surface. We were less lucky with the illumination conditions, which in the final location are very bad. So after the primary mission on batteries is over, we most likely won't have much of a chance to recharge the batteries. However this was always considered a ‘bonus’ – the solar cells were anyway too weak to perform many useful operations, even in best illumination conditions. We will learn more (and perhaps we'll have to correct the current understanding as I described above) once we get a final localisation with the Osiris pictures.
PF: After the end of the Philae relay phase, Rosetta will stay in a 20 km orbit for several weeks. Then, depending on the comet activity, it will increase the distance but continue following the comet in its journey towards the Sun. After perihelion in August 2015, we will try to go closer again as the activity decreases.
GS: What will we learn from the science?
PF: Well, it's not to me, the bus driver, to say what my passengers are up to! However in general Rosetta is about learning everything about comets. And since these objects are among the oldest objects in the solar system, the aim is to attempt to understand from what type of material the solar system was made, how it evolved, and how the planets formed. Also, are comets the source of the abundant water on the surface of the Earth, and maybe of life?
GS: ESA has truely 'come of age' with missions like the Titan landing in 2005, and now the Rosetta mission, which I rate as one of the most complex robotic missions. Why, in your view, does ESA not get the same kudos as NASA in the general sphere of astronautics? Is it a deficit in the ESA PR effort?
PF: I think Rosetta has changed this situation in the past few days. The resonance around the world, even in the US, was incredible.
GS: What does current analysis tell us about what actually happened at touch-down?
PF: We are learning more and more about this. We also have incredible pictures that show the lander flying over the surface after having touched down once! We are also close to finding the exact place Philae ended up after its two long jumps (it is very far from the initial touch down point!). What happened is simply that the harpoons did not fire at all. So Philae bounced, keeping its attitude since the flywheel was still rotating. However the first touch down on the surface automatically stopped powering the flywheel, which slowly reduced its rotation, thereby transferring angular momentum to the Lander body, which started rotating. After the first long jump that lasted almost 2 hours, Philae touched down once more and bounced again, but this time for a very short jump of 6-7 minutes, before stopping in its final position, most likely leaning against a rock or the wall of a crater. Fortunately we never lost the signal, and we had all the 5 foreseen contacts during the 2.5 days on the surface. We were less lucky with the illumination conditions, which in the final location are very bad. So after the primary mission on batteries is over, we most likely won't have much of a chance to recharge the batteries. However this was always considered a ‘bonus’ – the solar cells were anyway too weak to perform many useful operations, even in best illumination conditions. We will learn more (and perhaps we'll have to correct the current understanding as I described above) once we get a final localisation with the Osiris pictures.
Philae's final resting place - adjacent to a boulder or cliff. GS: Finally Paolo, will you see more of your family and friends post-landing?
PF: While Philae was active on the comet I spent most of the time in ESOC. I was with Philae on the comet, and slept 10 hours in 3 days. My family and children saw me only in TV interviews. Now that Philae sleeps I am at home, answering congratulation messages and catching up with work that last week had to wait for this historical phase to finish. So again not much for my family in this period I’m afraid. My team is still very busy with the continuation of the Rosetta mission, which is challenging enough even without having to take care of Philae in parallel. But I hope they can get some rest and family time over Christmas!
PF: While Philae was active on the comet I spent most of the time in ESOC. I was with Philae on the comet, and slept 10 hours in 3 days. My family and children saw me only in TV interviews. Now that Philae sleeps I am at home, answering congratulation messages and catching up with work that last week had to wait for this historical phase to finish. So again not much for my family in this period I’m afraid. My team is still very busy with the continuation of the Rosetta mission, which is challenging enough even without having to take care of Philae in parallel. But I hope they can get some rest and family time over Christmas!
RSS Feed
