Andy Pye revisits the Beagle 2 Mars landing to look at analysis performed to gain further insight into the space programme failure mechanisms.
In an article run in Environmental Engineering in February this year (The Beagle Has Landed), Andy Pye spoke to Professor Steve Burnage, chief engineer for Beagle 2’s Spin-Up Eject Mechanism (SUEM) about the testing processes which ultimately proved that it did perform its objective of getting the probe to Mars, after all. Now, a further study finds that the Mars lander deployed at least three or all four of its solar panels.
Beagle 2 was part of the ESA Mars Express Mission launched in June 2003. Mars Express is still orbiting Mars and returning scientific data on the planet. Beagle 2 was successfully ejected from ESA’s Mars Express spacecraft on 19 December 2003 but failed to send a signal on Christmas Day – its scheduled landing day on Mars. It was presumed lost until over a decade later when the mystery of what happened to the mission was solved through images taken by NASA’s Mars Reconnaissance Orbiter (MRO).
On landing, the probe was supposed to open like a clamshell and deploy solar panels and instruments that would look for signs of life on Mars, but nothing more was heard from Beagle 2 after it was ejected from its mothership and it lost contact..until early in 2015.
Now, thanks to an innovative research technique, a collaboration between De Montfort University and the University of Leicester has moved one step closer to understanding exactly what happened to the ill-fated Mars Lander Beagle 2.
Researchers from De Montfort University and the University of Leicester have worked together to come up with a new way to detect the configuration of the lander – the concept of “reflection analysis” – matching simulated and real images of Beagle 2. A 3D modelling technique is based on simulating possible configurations of the lander on the surface and comparing the light of the Sun reflected by the simulated lander with the unprocessed images available from the HiRISE camera at a number of different sun angles.
It reveals for the first time that Beagle 2 deployed at least three, and possibly all four, of the solar panels it was supposed to after touching down on the planet’s surface. The finding will rewrite scientific knowledge about the stricken Lander – it was previously thought that perhaps only (as few as) two of the four solar panels had deployed.
Professor Mark Sims, former Beagle 2 Mission Manager and Professor of Astrobiology and Space Instrumentation at the University of Leicester, turned to a team at De Montfort University to realise his concept. Commercially available software used for 3D modelling, animation, visual effects and simulation design was adapted to enable this analysis.
“The De Montfort team were responsible for all the 3D simulation work to test the reflection analysis concept. In order to do this, our visualisation specialist Teodora Kuzmanova had to create a physically accurate 3D model of the Beagle 2 Mars Lander with surfaces that would accurately reflect virtual sunlight. The angle of the sun had to be simulated, along with the position of a virtual camera that could take pictures equivalent to NASA’s Reconnaissance Orbiter. Finally these images had to be pixelated to match the resolution of the Orbiter’s images,” says Nick Higgett, leader of the De Montfort University Simulation team.
Kuzmanova and Dr Eric Tatham used 3D software to model the scene in three dimensions, adjusting the position of the sun and the resting angle and orientation of the Beagle 2, unfolding the four solar panels at different angles taking in the illumination conditions on the planet until they found the best visual match to what the NASA original images showed. These simulations were then adjusted to reproduce the resolution and view point of the NASA spacecraft.
Higgett explains: “The visual comparison between the real and simulated images could then begin to identify which landing configuration (1, 2, 3 or 4 deployed solar panels) was the best fit. This was originally a proof of principle project. However, we have gone way beyond this original plan to reach the exciting conclusion that Beagle 2 did not crash but landed and probably deployed most of its panels. Hopefully these results help to solve a long held mystery and will benefit any future missions to Mars.”
Sims adds: “Although the concept of the “reflection analysis” was mine, I didn’t know it would work. Thanks to the effort of the team at De Montfort University, they proved that this concept could work and we have gathered more information on the failure of Beagle 2 to communicate and we are one step closer to knowing what happened. In reality, we may of course never know exactly what caused its failure to communicate after what has been confirmed as a successful landing, which was a fantastic achievement by the Beagle 2 team. The work shows frustratingly that Beagle 2 came so close to working as intended on Mars.”
This work confirms that antenna transmission would probably have been hampered by one of the panels failing to unfold correctly, confirming the previously supposed theory. It complements other techniques, such as super-resolution imaging, as conducted by Professor Jan-Peter Muller and his team at University College London and announced in April 2016. Higgett says it is as close to a definitive explanation as would be possible without landing on the planet itself.
There are a number of issues associated with the concept. Firstly, a large number of possible combinations for the lander exist, initial modelling, some ray tracing and comparison by eye indicated that only a few combinations of orientation would give the observed changes in the real images with the sun angle. This then allowed modelling to concentrate on the number of panels deployed, the “exact” angle of orientation and tilt of base, lid and panels.
The second issue is where you place the boundaries of the pixels, here a number of different starting positions were taken and again a best match by eye approach was taken to eliminate poor matches before using mathematical techniques.
Thirdly De-Montfort have adapted commercial software to do the simulations, correcting for Martian illumination and using measured or estimated reflectivities for surfaces. An exact simulation would require bespoke software however it is believed that the inaccuracies introduced here are relatively small. However tilts and angles derived from the method should be treated as approximate.
Despite the above issues, the researchers believe that the concept is valid and provides important information on the configuration of Beagle 2 on the surface of Mars.
The “reflection analysis” technique used for this research could find applications in other fields where an illumination source is present and the target has a limited set of configurations and is highly reflective in nature.