The Mars Exploration Rover team scored an unexpected interplanetary hole-in-one in 2004, when the airbag encased Opportunity rover came to rest inside a crater containing exposed bedrock. But putting serendipity aside, the rover’s projected landing area was an ellipse about 85 kilometers (53 miles) long and 11 kilometers (6.8 miles) wide within a fairly benign landscape. The landing sites for all of NASA’s Mars landers have, to this point, been mainly wide, flat plains chosen to maximize science return while minimizing the risk posed by hazards on the planet’s surface. Similar concerns guided the selection of landing sites for the Apollo missions to the moon.
Now a team at the Jet Propulsion Laboratory is developing an Entry, Descent and Landing, or EDL, system which could allow safe landings much closer to targets of scientific interest than previously possible. Such a system has the potential to open up vast areas of the planets that have previously been off limits to exploration.
The new precision EDL system uses images of the planet’s surface taken during descent to navigate precisely to a desired target on the ground. The Spirit and Opportunity rovers flew unguided entries, spinning like a rifle bullet for stability, with landing uncertainties mostly due to the density of the top of the Martian atmosphere, which can vary daily. Just before landing, the rovers used an imaging system to determine whether a last-minute thruster firing was required to reduce their horizontal velocity at touchdown, in order to avoid ripping open the airbags that cushioned their arrival. NASA’s Mars Science Laboratory rover, slated for a 2011 launch, will steer using a small amount of lift to compensate for the atmospheric density uncertainty, guiding its trajectory through the atmosphere and reducing its landing uncertainty to an area roughly 20 kilometers (12 miles) wide. The new precision EDL system will add the use of images in the final stages of descent to guide a future Mars lander to within perhaps a hundred meters (about 300 feet) of a target site, which is a level of precision no previous lander has achieved.
Thus, the landing system the JPL team has in mind is a potential game changer in two ways. First, it would allow a probe to land at a specific target of science interest no larger than a football field. Secondly, it would allow a safe landing in a desired target region containing hazards.
Aron Wolf is the principal investigator on this JPL Research & Technology Development project. “This approach should allow us to land safely in places with more varied or rougher terrain, like valleys or features that look like dry riverbeds,” he says. “It could also help enable a Mars sample return mission.”
Wolf explains that, to return a sample from Mars, you have to land, gather the sample, and then launch it from Mars back to Earth. This may have to be done in several steps: first, landing a rover to collect the sample, then landing a small rocket close to the rover so the rover can place the sample on the rocket for return to Earth. “Precision landing would be essential for placing the second lander close to the rover,” Wolf says.
Building a smarter lander
Wolf’s project will produce prototype algorithms for key aspects of precision landing: guidance during powered descent, image processing for navigation using surface features and filtering of multiple types of navigation data, including imaging and radar altimetry. As a capstone, the team will run complete EDL simulations that emulate flight hardware in a Mars-like environment – a critical step for testing the algorithms as a system and demonstrating that all the elements can work together to accomplish pinpoint landings.
One of the system’s key capabilities will be the ability to compare images taken by a spacecraft during its descent with existing maps to determine its position relative to the landing target. A lander utilizing this technique, called terrain-relative navigation, would match landmarks in its own descent images with surface albedo and digital elevation maps produced by previous missions.
A Matter of Time
Using images to navigate is a particular challenge for Mars landers because there is little time available to combine images with other navigation data to determine position and velocity – perhaps only 30 seconds to a minute. Accurate measurements must be obtained before the descent engine is fired, but an onboard camera cannot begin imaging until the parachute deploys and the heat shield is jettisoned. With these challenges in mind, the team is seeking to understand how many images are needed and how quickly their system will have to analyze them.
Even after it fires its rocket engine, a lander may have to correct its course to get to a preselected landing site. Every second burns precious fuel. So Wolf and his team must find ways to minimize the propellant needed to get to the target in order to maximize the mass of the payload that can be delivered to the planet’s surface.
In addition, the team must ensure that their software can run as fast as needed within the anticipated computing capabilities of future spacecraft. For current missions, ignition of the descent engine occurs at a fixed altitude and the engine’s operation is largely predetermined. A future lander performing autonomous navigation will need to quickly determine its optimal engine firing sequence with limited onboard computing power, and Wolf and colleagues propose to do this with a simple solution: do the number crunching before leaving Earth. Instead of forcing the onboard computer to expend precious resources during the hectic period of powered descent, the precision EDL system would compile its navigation inputs and look up the solutions in tables stored in memory. That way the spacecraft has only to plug in numbers to get the optimal navigation solutions it needs.
Currently in the second year of their three-year project, Wolf and his team aim to lay the groundwork for an EDL system that will be mature enough for implementation on a Mars mission in the near future. But while emphasis in the precision landing system’s development is on missions to the Red Planet, the technology could be employed by spacecraft making visits to lots of other worlds. According to Wolf, this new capability could represent “one small step for a machine, one giant leap for mankind.”
Other JPL team members on this Research & Technology Development project are Behcet Acikmese, J. "Bob" Balaram, Lars Blackmore, Jordi Casoliva, Yang Cheng, Martin Heyne, Mark Ivanov, Chris Lim, Nick Mastrodemos and Ryan Park.
Links
- Mars Technology Program: http://marstech.jpl.nasa.gov/index.cfm
- JPL Robotics: http://www-robotics.jpl.nasa.gov