Behind the Mars Exploration Rover Cameras
Black Magic or a Bag of Tricks?
In his office at NASA's Jet Propulsion Laboratory, Optics Engineer Larry Scherr sits before a buzzing computer, sketching geometrical shapes on a sheet of graph paper. He is calculating lens shapes to control the path of light rays.
Image Right: The Mars Exploration Rovers' panoramic cameras are capable of taking color images such as this one of Opportunity's lander, the Challenger Memorial Station, at Meridiani Planum, Mars. + Click for full image. Image courtesy: NASA/JPL/Cornell.
"It's puzzling," says Scherr, pointing to a series of half-oval shapes. "These are mirrors. I think I need four of them to be able to make this work."
Scherr and his colleagues are professional problem-solvers who receive proposals for amazing scientific instruments and try to make them a reality. Four years ago, Scherr was analyzing lenses for the Mars Exploration Rover cameras, now sending back a myriad of eye-popping images from the red planet.
"It's nice to work on something from concept to design, and talk with the people that are building the hardware," Scherr said. "It's through this kind of interaction -- being able to work through problems together -- that we made these cameras work and return images from Mars within a few years."
By "we," Scherr means the more than 100 engineers, contractors, graduate students and individuals from universities and research institutions that helped design, build, calibrate and test the Mars Exploration Rover cameras. Approximately 35 of those people were from JPL.
Conjuring Images, Step One
In 2000, Mars Exploration Rover scientists gave camera engineers at JPL a lengthy list of requirements for the cameras that would eventually play a key role in the rover's mission on Mars -- phase one of a long assembly line of work that produced 42 cameras: 20 on the twin rovers, 18 for test rovers on Earth and four spares. Each rover would contain two panoramic cameras, four hazard avoidance cameras, two navigation cameras, a microscopic imager camera, and a descent image motion estimation system camera, located on the lander.
"Time was so tight for this project," said JPL Camera Payload Element Manager Mark Schwochert. "We had to build a large number of cameras in just two years, so we tried to keep the design simple, yet robust."
"We had two classes of requirements, one for pure science, and the other for mission support," said Dr. Jim Bell, camera payload element lead scientist from Cornell University, Ithaca, N.Y.
A key science and engineering requirement called for two high-resolution, color cameras known as the panoramic cameras on each rover. These cameras, used together, would provide full-color, three dimensional, 360-degree panorama images of the terrain around each rover -- a very important aid to engineers who are driving the rovers.
"We set a benchmark with the camera by asking that they have resolution at least as good as the human eye with perfect vision," Bell said. "And of course, we wanted really stunning pictures."
To get these images, scientists choose which filters to take the image with out of 11 available color filters. These filters help them acquire and reconstruct a reasonably accurate color image and, more importantly, determine the composition of rocks and soil on the martian surface. An additional two filters are included to help photograph the Sun. This not only helps scientists determine the dust abundance in the atmosphere, but also allows each camera to serve as a "solar compass" to determine a rover's orientation.
"The rovers really are like robotic geologists," said JPL Senior Test Engineer Tom Elliot. "What the cameras are seeing on Mars is very similar to what we would see if we were there looking around."
The camera team included specialists from five major disciplines: optics, charge coupled devices, electronics, mechanicals, and assembly and testing.
Image Left: The lenses in the Mars Exploration Rover cameras range from 8 to 12 millimeters (0.3 to 0.5 of an inch) in diameter, about the size of a soft contact lens. + Click for full image. Image courtesy: NASA/JPL.
With valuable input from key contractors, the optics team at JPL designed, built and tested the tiny camera lenses, each custom-tailored to the function its camera was to perform. To meet their requirements, the team had to carefully consider using smaller consumer parts, some of which weren't capable of handling a journey to Mars.
"One of the stressful things about being an optical designer is that you have to be one of the first ones finished so other engineers in the mechanical and thermal areas may do their work," said Principal Optical Engineer Ed Hagerott. "All the people who worked on this project put in very long hours and did a super job of meeting the requirements, sometimes even ahead of schedule."
Step Two, Three and Voila! Images!
While the optics team developed the lenses, a second team designed, built and tested charge coupled devices, the "electronic film" which would convert the optical images (patterns of light, or an image as we would normally see it) into patterns of electronic signals. These signals would then be radioed to Earth and reconstructed into pictures.
Image Right: Charge coupled devices, or CCDs, resemble computer chips, although their function is very different. Rather than processing an electronic signal, as a computer chip does, a charge coupled detector transforms a pattern of light (an optical image) into a pattern of electrical charge (an electronic image). + Click for full image. Image courtesy: NASA/JPL.
Meanwhile, a third team developed the electronics that would operate the camera and process the charge coupled devices' electronic signal. This provides electronic images to the rover computer for transfer to Earth, often via spacecraft that are orbiting Mars.
"Acquiring the images using the rover cameras isn't the black magic that it seems," said Elliot. "The operations team has to decide which filters to use, command the camera, and then process the resulting image after it has been transferred to Earth. In fact, the rover cameras, in many ways, are less complicated than consumer digital cameras."
Practice Makes Perfect
"By the time the mission team got the cameras, it was like buying a new car with 25,000 miles already on it," Schwochert said. Prior to delivery, the camera team performed a rigorous series of tests and reviews, referencing designs from Sojourner, Viking and other missions for ideas and problems that they might encounter, such as whether dust would accumulate on the lenses.
"Different missions have different ways of fulfilling their requirements, and it was helpful for us to look at what was successful and where problems were encountered in the past," Elliott said. "Of course, we had our own bag of tricks."
The team conducted camera and rover tests in vacuum chambers and in JPL's In Situ Instruments Laboratory, a building that contains a patch of realistic "Mars" terrain. Camera testing included operation at temperatures as low as -120 degrees Celsius (-184 degrees Fahrenheit), even though the rover cameras on Mars are typically operated at temperatures above -45 degrees Celsius (-49 degrees Fahrenheit).
"After days of electrical testing and retesting under the most rigorous conditions that I could think of, I knew that the camera electronics design was solid, but in this business there is always room for more testing," said Lead Camera Electronics Engineer Arsham Dingizian. "I wanted to be prepared for the worst because no design is perfect. To my surprise, although they didn't work properly at -120 degrees Celsius, the cameras electronics survived and worked perfectly at above -55 degrees Celsius (-67 degrees Fahrenheit). This proved to me that the camera electronics design not only met its specifications but also exceeded them. At that moment I knew that the camera electronics could handle almost anything Mars could come up with."
After the cameras were built and thoroughly tested to ensure mission success, the camera team turned them over to the Assembly, Test, Launch and Operations team, who placed them on the rovers. Members of the camera team have since moved onto other projects, although a handful of them still work daily with the Mars Exploration Rovers.
"There was so much riding on these missions, I found myself waiting, like everyone else, into the early hours of the morning until the first images came down," said Elliott, who no longer works with the rovers but is now helping develop the navigation camera on the 2005 Mars Reconnaissance Orbiter and the 2009 Space Interferometry Mission.
NASA's Jet Propulsion Laboratory