Philae (spacecraft)

• 1. Introduction

  ›

  Philae touched down on a comet on the 12th of November 2014, then bounced twice before coming to rest on its side. This was the first attempt to land softly on a comet nucleus. The European Space Agency's robotic lander had traveled with the Rosetta spacecraft for ten years and eight months to reach it. Its harpoons failed to fire. Its anchoring thruster never fired either. Yet Philae still managed to send back the first images ever taken from a comet's surface. It also gathered data on what a comet is actually made of. How did a lander built to anchor itself end up rolling across an alien landscape instead? What did it manage to learn in the short window of power it had? And how, months later, did anyone even find it again on a body drifting through space?

• 2. A Decade in Transit

  ›

  An Ariane 5G+ rocket carried Rosetta and Philae away from French Guiana on the 2nd of March 2004, at 07:17 UTC. The pair then spent 3,907 days, close to 10.7 years, crossing space toward comet Churyumov-Gerasimenko. Philae's job once it arrived was simple to state: land on the comet, anchor itself, and transmit data about what the comet was made of. Unlike the Deep Impact probe, which by design struck comet Tempel 1's nucleus on the 4th of July 2005, Philae was never built to crash.

  On the 25th of February 2007, Rosetta swept past Mars, and several of Philae's instruments ran on their own for the first time. CIVA, one of the camera systems, captured images while Rosetta's own instruments sat powered down for the flyby. ROMAP, meanwhile, measured the magnetic environment around Mars. Most of Philae's other instruments needed direct contact with a surface to work, so they stayed offline throughout the flyby. Mission planners at the time floated an optimistic guess for how long Philae might function after touchdown: "four to five months."

  Rosetta's original destination was comet 46P/Wirtanen, not Churyumov-Gerasimenko. A failure in an earlier Ariane 5 launch closed the window to reach Wirtanen, forcing mission planners to retarget the spacecraft. Churyumov-Gerasimenko's greater mass meant a faster impact velocity, so engineers had to strengthen Philae's landing gear to compensate.

  The lander's own name reaches back further than the mission itself, to the Philae obelisk in Egypt. That obelisk carries a bilingual inscription, and it was used alongside the Rosetta Stone to decipher Egyptian hieroglyphs. Once under way, Philae was monitored from DLR's Lander Control Center in Cologne, Germany, with support from CNES's SONC facility in Toulouse, France.

  By the 6th of August 2014, Rosetta had rendezvoused with the comet, and Philae's actual landing day was still more than three months away.

• 3. Three Touches Down

  ›

  Mission planners announced "Site J" on the comet's smaller lobe as the landing target on the 15th of September 2014. After a public contest that October, Site J was renamed Agilkia, honoring Agilkia Island.

  A series of four go or no-go checks ran on the 11th and the 12th of November 2014. One of the final tests revealed that Philae's cold-gas thruster was not working correctly, but since it could not be repaired, controllers gave the go-ahead anyway. Philae detached from Rosetta on the 12th of November 2014, at 08:35 UTC SCET.

  Philae's descent plan called for a drop from an orbital altitude of 22.5 kilometres along a ballistic path, aiming for a touchdown speed of around 1 metre per second. Legs were built to absorb that impact without a bounce, since engineers had estimated the comet's own escape velocity at only about 1 metre per second, and the force of contact was meant to drive ice screws into the ground. A harpoon was then supposed to fire into the surface at 70 metres per second to anchor the lander, backed by a thruster meant to press it down and offset the harpoon's recoil. Neither the harpoon nor the thruster fired as planned.

  Earth's tracking stations received Philae's landing signal at 16:03 UTC, after a 28-minute delay across space. Unknown to the science team at the time, the lander had already bounced. Philae first touched the comet's surface at 15:34:04 UTC SCET, rebounding at 38 centimetres per second and rising roughly a kilometre back off the surface. Had it left the comet at more than about 44 centimetres per second, it would have escaped the comet's gravity entirely.

  Philae's reaction wheel powered off automatically once touchdown was detected, and the stored momentum spun the lander once every 13 seconds. At 16:20 UTC SCET, during that first bounce, the lander is thought to have struck a surface prominence, which slowed its spin to once every 24 seconds and sent it tumbling. Philae touched the surface a second time at 17:25:26 UTC SCET, rebounding gently at 3 centimetres per second, before coming to a final stop at 17:31:17 UTC SCET. It ended up in rough terrain, apparently in the shadow of a cliff or crater wall, tilted at around 30 degrees but otherwise undamaged.

  Telemetry later showed the initial impact had actually been softer than engineers expected. The harpoon's propulsion charge held just 0.3 grams of nitrocellulose, a compound that Copenhagen Suborbitals had already shown, in 2013, to be unreliable in a vacuum.

  Exactly where the lander had come to rest would not be confirmed by direct imaging for almost two years.

• 4. Racing the Clock

  ›

  Philae's primary battery was designed to power its instruments for about 60 hours. ESA had expected a secondary rechargeable battery to be topped up by the lander's solar panels, but the actual landing site received only about 90 minutes of sunlight during each 12.4-hour comet day, far too little to sustain operations.

  By the morning of the 14th of November 2014, mission control estimated the remaining battery charge would only last through the rest of that day. Controllers first pulled data from instruments that needed no moving parts, covering roughly 80 percent of the planned initial science program. They then commanded the MUPUS soil penetrator and the SD2 drill to deploy, and MUPUS, COSAC, and Ptolemy all returned data, along with a final batch of readings from CONSERT.

  During that evening's transmission window, ground controllers raised Philae by 4 centimetres and rotated its body 35 degrees, angling its largest solar panel toward whatever sunlight might come. Electrical power then dropped away quickly, forcing every instrument to shut down, and the downlink slowed to nothing. Contact was lost on the 15th of November at 00:36 UTC, and Philae slipped into a hibernation known as safe mode as its batteries ran down.

  DLR lander manager Stephan Ulamec summed up the moment: "Prior to falling silent, the lander was able to transmit all science data gathered during the First Science Sequence... This machine performed magnificently under tough conditions, and we can be fully proud of the incredible scientific success Philae has delivered."

  Mission controllers still held out hope that more sunlight on the solar panels might eventually be enough to reboot the lander.

• 5. Ten Instruments

  ›

  Philae's science payload comprised ten instruments weighing 26.7 kilograms altogether, just over a quarter of the lander's total mass of about 100 kilograms. The main body was built from carbon fibre, shaped into a stability plate, an instrument platform, and a hexagonal frame holding it together, with solar cells covering its exterior.

  APXS, the Alpha Particle X-ray Spectrometer, detected alpha particles and X-rays to reveal the elemental makeup of the comet's surface, building on an earlier version flown on the Mars Pathfinder mission. CIVA, the Comet Nucleus Infrared and Visible Analyser, combined seven identical cameras arranged around the lander at 60-degree intervals. Five of the cameras took single images, while two worked together as a stereo pair, each with a 1024 by 1024 pixel sensor. A separate microscope and infrared spectrometer, mounted at the base of the lander, studied the texture and reflectivity of collected samples.

  CONSERT, the Comet Nucleus Sounding Experiment by Radiowave Transmission, sent a radio signal from Rosetta through the comet's nucleus to a receiver on Philae, mapping its internal structure. COSAC, the Cometary Sampling and Composition instrument, paired a gas chromatograph with a time-of-flight mass spectrometer to analyse soil samples for volatile compounds. MUPUS, the Multi-Purpose Sensors instrument, measured the density, temperature, and mechanical strength of the surface and subsurface.

  Ptolemy measured the stable isotope ratios of key volatiles in the comet's nucleus, with parts built by the Special Techniques Group at UKAEA. ROLIS, the Rosetta Lander Imaging System, used a 1024 by 1024 pixel CCD camera to capture high-resolution images during descent and of areas other instruments sampled. ROMAP, the Rosetta Lander Magnetometer and Plasma Monitor, studied the comet's magnetic field and how it interacted with the solar wind.

  SD2, the Sampling, Drilling and Distribution system, drove a steel and titanium drill up to 230 millimetres into the comet, feeding samples to 26 platinum ovens, ten heated to 180 degrees Celsius and 16 to 800 degrees Celsius. Development of SD2 was led by the Italian Space Agency, with contractor Tecnospazio S.p.A., along with Tecnomare S.p.A., Media Lario, and Dallara; its principal investigator was Amalia Ercoli-Finzi of Politecnico di Milano. SESAME's three instruments rounded out the payload: CASSE measured how sound moved through the surface, the Permittivity Probe studied its electrical properties, and the Dust Impact Monitor tracked dust falling back down.

  Power came from two batteries: a non-rechargeable 1000 watt-hour primary cell and a 140 watt-hour secondary cell meant to recharge from the panels once the primary ran out. Those solar panels covered 2.2 square metres and were designed to deliver up to 32 watts at a distance of 3 astronomical units from the Sun. Communications relied on Rosetta as a relay station, which reduced how much electrical power Philae itself needed to spend on transmission.

  The mission's surface operations were planned to last at least one week, with an extended run of months considered possible if conditions allowed.

• 6. What the Comet Revealed

  ›

  SESAME's readings from Philae's first touchdown site upended expectations. Rather than the soft, fluffy material scientists had anticipated, the ground held a large amount of water ice beneath about 25 centimetres of granular material. That ice proved mechanically strong, while the surrounding area showed little cometary activity.

  At the final landing site, though, MUPUS could not hammer far into the surface even as its driving power was gradually increased. That ground had the consistency of solid ice or pumice.

  COSAC detected molecules containing carbon and hydrogen in the thin atmosphere around the comet. It could not analyse soil elements directly, since the lander was unable to drill into the surface, likely because the ice there was too hard. SD2's drill went through every step needed to deliver a sample to COSAC, but no material actually entered its ovens.

  When Philae first touched the surface, COSAC sampled material disturbed at the bottom of the vehicle while Ptolemy measured material at the top. Between them, the instruments detected sixteen organic compounds. Four of those, acetamide, acetone, methyl isocyanate, and propionaldehyde, had never before been found on a comet.

  In October 2020, the journal Nature published findings on what Philae had actually discovered during its working hours on the surface. Researchers reported low-strength, primitive ice inside cometary boulders, including water ice dating back to the comet's formation an estimated 4.5 billion years earlier. This turned up mainly at Philae's second touchdown site, where the lander made four distinct contacts across two adjoining boulders, drilling 25 centimetres into the ice there.

  None of this analysis was straightforward. Confirming exactly where on the comet these boulders sat depended on pinpointing Philae's own resting place, a question that would take Rosetta years to answer.

• 7. Lost and Found

  ›

  On the 13th of June 2015, at 20:28 UTC, ground controllers received an 85-second transmission relayed through Rosetta: Philae was awake, its batteries recharged enough to leave safe mode. The lander reported it had actually been active before that date, though it had been unable to reach Rosetta until then, and it was running on 24 watts of power at a temperature of minus 35 degrees Celsius.

  A second contact followed on the 19th of June 2015, with signals received on the ground at 13:37 UTC and again at 13:54 UTC, each lasting about two minutes. By the 26th of June, controllers counted seven intermittent contacts in total. Each day offered two possible windows, but their length and clarity depended on Philae's antenna orientation, Rosetta's position, and whether the comet's rotation had turned the lander toward the sun. ESA controllers kept aiming for one stable link lasting at least 50 minutes.

  Had Philae landed at its planned site of Agilkia back in November 2014, engineers believed solar heating would probably have ended the mission by March 2015 anyway. Drilling into the surface to determine the comet's chemical makeup remained Philae's key unfinished experiment.

  On the 5th of July 2015, controllers sent commands to power up the CONSERT radar, and confirmation that it had worked did not arrive until the 9th of July, when the lander transmitted measurement data. That transmission turned out to be the last one. Rosetta, forced to keep a safer distance as the comet grew more active, could not re-establish a stable connection, and by January 2016 controllers accepted that further contact was unlikely. On the 27th of July 2016, at 09:00 UTC, ESA switched off Rosetta's Electrical Support System Processor Unit, closing the door on any further contact with Philae.

  Throughout the mission, engineers kept comparing pictures taken before and after the landing, searching for a patch of ground with Philae's telltale reflectivity. On the 2nd of September 2016, Rosetta's narrow-angle camera finally confirmed it, in an image taken from 2.7 kilometres away that clearly showed the lander. Philae was lying on its side, wedged into a dark crevice, which explained why it had never had enough sunlight or a clear line of communication. That single image gave scientists the context they needed to make full sense of Philae's two days of active science on the surface.

  Less than a month later, on the 30th of September 2016, Rosetta itself ended its mission by descending into the comet's Ma'at region.

  Building and flying Philae had already drawn engineers and institutions from more than a dozen countries into the project long before anyone knew whether the landing would work.

• 8. Built by the World

  ›

  Austria's Space Research Institute built Philae's anchor and two of the sensors inside MUPUS, fitted directly into the anchor's tips. Belgium's Institute for Space Aeronomy helped build one of the sensors in Rosetta's ROSINA instrument and, together with the Royal Observatory of Belgium, tracked space weather conditions, particularly solar proton events, ahead of the landing. Canada contributed through SED Systems in Saskatoon, which built three ground stations for communicating with Rosetta, and through ADGA-RHEA Group of Ottawa, which supplied the software managing command sequences. Finland's Meteorological Institute provided memory for the lander's Command, Data and Management System and for its Permittivity Probe.

  France's national space agency, working with institutes including IAS, SA, LPG, and LISA, handled the lander's overall engineering, radio communications, battery assembly, CONSERT, CIVA, and ground operations. Germany's DLR supplied the structure, thermal systems, flywheel, descent system, ROLIS camera, and the SESAME acoustic and seismic hardware, and managed the project overall. The University of Münster built MUPUS, which had been designed at Poland's Space Research Centre, while the Braunschweig University of Technology built ROMAP. The Max Planck Institute for Solar System Research handled payload engineering, the landing gear, the anchoring harpoon, the central computer, COSAC, and APXS.

  Hungary's Wigner Research Centre for Physics designed the Command and Data Management Subsystem, while Budapest University of Technology and Economics designed the power subsystem controlling battery charging. In Ireland, Captec Ltd. of Malahide validated mission-critical software and built the communications interface, while Space Technology Ireland at Maynooth University built the Electrical Support System Processor Unit that relayed commands and data. Italy's Space Agency developed the SD2 drill and the photovoltaic assembly, and Italian Alenia Space handled assembly, integration, testing, and the probe's radio transponder.

  Moog Bradford in the Netherlands built the Active Descent System that guided Philae down to the surface, working with Switzerland's Bleuler-Baumer Mechanik, while Switzerland's Centre for Electronics and Microtechnology separately developed CIVA. Spain's GMV calculated the lighting and visibility conditions needed to choose a landing point, with further contributions from INTA and the Spanish division of Airbus Defence and Space. In the United Kingdom, the Open University and the Rutherford Appleton Laboratory built Ptolemy, Surrey Satellite Technology built the momentum wheel that stabilised Philae during descent, and manufacturer e2v supplied its camera systems.

  The landing itself played out heavily on social media, complete with an official Twitter account written in Philae's own voice. The hashtag #CometLanding took off worldwide, amplified by a livestream from the control centres and by watch events held both officially and informally around the globe. Several of Philae's individual instruments even had their own Twitter accounts to announce results.

  Composer Vangelis wrote the music for a trio of videos ESA released to mark the first attempted soft landing on a comet. Google marked the occasion too, with a Doodle of Philae on its homepage on the 12th of November 2014, followed by another appearance in Google's New Year's Eve Doodle on the 31st of December that year. Webcomic author Randall Munroe spent the day of the landing writing a live-updating strip about it on his site, xkcd.