Researchers with the ExoMars mission are pointing to a potential computing glitch as the cause of last week's crash of the Schiaparelli lander. The challenge now will be to isolate and correct the error in hopes of preventing a repeat in 2020, when mission planners aim to land a much larger rover on the Red Planet. Illustration showing the moment the Schiaparelli lander was to jettison its back shell and parachute and ignite its descent engines. A preliminary analysis suggests this stage happened prematurely, likely the result of a software error. (Image: ESA/ATG Medialab)
Late last week, NASA released a grim photograph showing what appears to be the crash site of the doomed Schiaparelli lander and its discarded parachute. The lander, if there was ever any doubt, is completely toast, a splotch of burnt and twisted metal on the Martian surface. So instead of proudly carting out a new rover, ExoMars planners with the ESA and the Russian space agency Roscosmos are now having to figure out what the hell went wrong.
Unlike the doomed Beagle 2 mission that was lost in 2003, Schiaparelli transmitted its status data to its mother ship — the Trace Gas Orbiter — during its descent. As reported in Nature News, an early look at the data points to a series of cascading software errors as the reason for the botched landing.
By all accounts the descent started well, with the lander decelerating rapidly as it brushed up against the Martian atmosphere, eventually deploying its parachute as planned. But things began to go squirrely just prior to the five-minute mark of the planned six-minute descent.
For reasons that are still a mystery, the lander ejected both its heat shield and parachute way ahead of schedule. Schiaparelli then engaged its thrusters for a painfully brief three-second burst — a procedure that was supposed to last for 30 seconds once the lander was just a couple of metres off the ground. The lander's onboard computer, it would appear, seems to have thought it was close to the surface. Indeed, Schiaparelli even took the time to switch on some of its instruments, including tools to record the planet's weather and electrical field.
The sad reality is that Schiaparelli was still somewhere between 2km to 4km above the surface when this happened, falling at a rate of about 300km/h. It struck the ground with tremendous force, resulting in an explosion — and a brand new surface feature.
ESA's head of solar and planetary missions, Andrea Accomazzo, suspects a flaw in Schiaparelli's software, or a problem in integrating the data coming from different sensors. Some kind of glitch misinformed the lander about its position in time and space, causing it to execute landing procedures as if it were at a much lower altitude.
If confirmed, this would actually be good news, as software issues are much easier to correct than hardware problems. Researchers on the ExoMars team are confident in the integrity of Schiaparelli's hardware, and they're now hoping to replicate the software error using a simulation.
If and when the glitch is detected, a fix will have to be designed, implemented and tested. ExoMars planners don't have much time, as the second and most prominent part of the mission is scheduled for 2020. This first phase was meant as a kind of test-run in preparation for the landing of the larger Russian ExoMars rover. Let's hope history doesn't repeat itself in four years time.