Each rover we send to Mars is more autonomous in its navigation than the previous rovers. Curiosity is more autonomous than Spirit and Opportunity, which were more autonomous than Sojourner.
This is due to the fact of better code writing by the engineers, better computers (and extra computers) and mostly by the many extra cameras couples with the needed processing power in the on board computers.
Like Spirit and Opportunity before her the Curiosity rover has two sets of engineering cameras: navigation cameras (Navcams) up high on the rover and hazard-avoidance cameras (Hazcams) down low.
All these cameras will help operational decisions both by Curiosity’s onboard autonomy software and by the rover team on Earth. Information from these cameras is used for autonomous navigation, engineers’ calculations for manoeuvring the robotic arm and scientists’ decisions about pointing the remote-sensing science instruments.
In 2010 (6 years into her extended mission) Opportunity was upgraded with software to allow her to make autonomous decisions on where to drive. This new system is called Autonomous Exploration for Gathering Increased Science, or AEGIS. Without autonomous driving software the rover has to send observations to Earth, where engineers have to judge it, then they will request follow-up observations, before they can program the rover to drive. Autonomous driving means a far greater efficiency with the time of the rover spend on Mars, and with its team of operators on Earth.
The cameras on Curiosity are virtually identical to the engineering cameras on Spirit and Opportunity, although Curiosity has extra cameras and slightly more powerful heaters for the cameras. Curiosity’s Hazcams have one-time-removable lens covers to shield them from potential dust raised during the rover’s landing. Pyrotechnic devices will remove the lens caps after landing. The Navcams have no lens covers as they are protected during the landing as the remote sensing mast is in a stowed position for the landing.
Curiosity has a total of 12 engineering cameras, each weighing about 250 grams (9 ounces).
two round camera eyes of the navcams can be seen here next tot he square openings of the Mastcam 34 (left) and the MastCam 100 (right). In this image it is also very clear that the MastCams can only take square images of 1200 x 1200 pixels as their aperture is square and not rectangular.
Curiosity’s pair of Navcams are installed next to the science payload’s Mast Camera on the remote-sensing mast.
The NavCams are capable of imaging 3D images. Curiosity carries two stereo pairs of Navcams, one pair each connected to the rover’s two redundant computers. The Navcams that are controlled by and feed imagery to the “A” computer are mounted directly above the ones linked to the “B” computer. That puts the “A” pair about 1.99 meters (6.5 feet) above the ground when the rover is on hard, flat terrain and the cameras are pointed straight out (slightly lower when pointed downward). The other pair of NavCams are 5 centimeters (2 inches) lower on the mast, 1.94 centimeters off the ground. Compared to the MastCam on both mars exploration rovers that is 40-45 centimeters higher. The left and right cameras in each pair are about 42 centimeters (16.5 inches) apart, about twice further apart than the navigation cameras on Spirit and Opportunity.
Each of the Navcams captures a square field of view 45 degrees wide and tall, comparable to the field of view of a 37-millimeter-focal-length lens on a 35-millimeter, single-lens reflex camera. The lens focuses the image onto a 1,024-pixel-by-1,024-pixel area of a charge-coupled device (CCD) detector. The Navcams and the Hazcams both generate greyscale images that cover red wavelengths centered at about 650 nanometers. This yields a resolution of 0.82 milliradians per pixel — for example, 0.8 inch (2 centimeters) per pixel at a distance of 82 feet (25 meters), enough to resolve a golf ball at that distance as a circle about two Navcam pixels wide. The depth of field achieved by the fixed-aperture f/12 Navcams keeps anything in focus from a distance of about 20 inches (0.5 meter) to infinity.
Curiosity has four pairs of Hazcams: two redundant pairs on the front of the chassis and two redundant pairs on the rear. The rover can drive backwards as well as forward. This made it important that both the front and rear Hazcams can be used for detecting potential obstacles in the rover’s driving direction. The front Hazcams also provide three-dimensional (3D) information for planning motions of the rover’s robotic arm, such as positioning of the drill or scoop for collecting samples.
Each Hazcam has a fisheye lens providing a square field of view 124 degrees wide and tall. The Hazcam can focus from 10 centimeters (about 4 inches) to infinity. Resolution of the Hazcams is 2.1 milliradians per pixel on the same type of detector as in the Navcams. At a distance of 10 meters (33 feet), the Hazcam resolution is 2 centimeters (0.8 inch) per pixel. A golf ball at 10 meters distance would be approximately two Hazcam pixels across.
The redundant pairs of Hazcams are mounted side by side. On the front, each stereo pair has a baseline of 16.6 centimeters (6.54 inches) between the center of its left eye and center of its right eye. Both the pair linked to the rover’s “A” computer and the pair linked to the “B” computer are mounted near the bottom center of the front face of the body, about 68 centimeters (27 inches) above ground level.
The rear stereo pairs each have the same baseline as the front and read hazard-identification cameras on Spirit and Opportunity, a baseline of 10 centimeters (3.9 inches). The rear “A” pair on Curiosity is toward the right hand side of the rear face of the vehicle (on the left if a viewer were standing behind the rover and looking toward it).The rear “B” pair is on the right, or toward left hand side. Both pairs are about 78 centimeters (31 inches) above ground level.
Different navigation modes for rover drives use images from the engineering cameras in different ways. Techniques include “blind” driving, hazard avoidance and visual odometry. The set of commands developed by rover planners for a single day’s drive may include a combination of these modes.
When using the blind-drive mode, rover planners have sufficient local imaging from the engineering cameras or Mast Camera to determine that a safe path exists, free of obstacles or hazards. They command the rover to drive a certain distance in a certain direction. In a blind drive, the rover’s computer calculates distance solely from wheel rotation; one full turn of a wheel with no slippage is nearly 63 centimeters (25 inches) of driving.
When the rover planners cannot determine that a path is free of obstacles, they can command driving that uses hazard avoidance. Besides using this hazard avoidance in rougher terrain, they might use it for an additional segment of driving beyond a blind drive on the same day. Hazard avoidance requires the rover to stop frequently to acquire new stereo imaging in the drive direction with the engineering cameras and analyze the images for potential hazards.
The rover makes decisions based on its analysis of the three-dimensional information provided by the stereo imaging. Rover planners set variables such as how frequently to stop and check, which cameras to use, and what type of decision the rover makes in response to a hazard detection (whether to choose a path around it or stop driving for the day).
Hazard avoidance can be supplemented with visual odometry. Visual odometry uses Navcam images made with the cameras pointed to the side of the route being driven. By pausing at intervals during the drive to take these images, the rover can compare the before-and after situation for each segment of the drive. It can recognize features in the images and calculate how far it has actually traveled during the intervening drive segment. Any difference between that distance and the distance indicated from wheel rotation is an indication that the wheels are slipping against the ground. In the set of driving commands for that day, rover planners can set the intervals at which the rover pauses for visual-odometry checks, based on the type of terrain being traversed. A slip limit can be set for a stretch of driving, so that if the rover calculates that it is slipping in excess of that amount, it will stop driving for the day. The mode of using visual odometry checks at intervals several times the rover’s own length is called slip-check, to differentiate it from full-time visual odometry with sideways-looking stops as frequently as a fraction of the rover’s length. Navigation modes differ significantly in the fraction of time spent with wheels in motion versus stopped for imaging and analysis of the images.
Curiosity can incorporate other safety features in each drive, such as tilt limits. The rover has gyroscopes to determine the tilt. These gyroscopes are part of the inertial measurement unit. The gyroscopes measure the changes in the tilt of the rover. It is important to know this before the robotic arm and the science instruments are set to work. It is also used before pointing the high-gain antenna.
Rovers can drive forward or backward. This 3D image of Opportunity driving backward in 2011 shows very distinctive patterns of wheeltracks. This is due to the fact that driving backward an antenna blocks her view partially. Opportunity stops every 1.2 meters (4 feet) to check her patch. She pivots to be able to see her path and this is visible in the driving tracks she lieves behind.
Besides having extra pairs of Hazcams and navcams Curiosity also has redundant main computers, or rover compute elements. Of this “A” and “B” pair, it uses one at a time, with the spare held in cold backup. Thus, at a given time, the rover is operating from either its “A” side or its “B” side. Most rover devices can be controlled by either side; a few components, such as the navigation camera, have side-specific redundancy themselves. The computer inside the rover — whichever side is active — also serves as the main computer for the rest of the Mars Science Laboratory spacecraft during the flight from Earth and arrival at Mars.
Having two computers is also helpful for the planning of the landing. In case the active computer resets for any reason during the critical minutes of entry, descent and landing, a software feature called “second chance” has been designed to enable the other side to promptly take control, and in most cases, finish the landing with a bare-bones version of entry, descent and landing instructions.
Each rover computer element contains a radiation-hardened central processor with PowerPC 750 architecture: a BAE RAD 750. This processor operates at up to 200 megahertz speed, compared with 20 megahertz speed of the single RAD6000 central processor in each of theMars rovers Spirit and Opportunity. Each of Curiosity’s redundant computers has 2 gigabytes of flash memory (about eight times as much as Spirit or Opportunity), 256 megabytes of dynamic random access memory and 256 kilobytes of electrically erasable programmable read-only memory.
The Mars Science Laboratory flight software monitors the status and health of the spacecraft during all phases of the mission, checks for the presence of commands to execute, performs communication functions and controls spacecraft activities.
The spacecraft was launched with software adequate to serve for the landing and for operations on the surface of Mars, as well as during the flight from Earth to Mars.
The months after launch were used, as planned, to develop and test improved flight software versions. One upgraded version was sent to the spacecraft in May 2012 and installed onto its computers in May and June. This version includes improvements for entry, descent and landing.
Another was sent to the spacecraft in June and will be installed on the rover’s computers a few days after landing, with improvements for driving the rover and using its robotic arm.