AI researchers within NASA are now focusing on deploying intelligent software directly on the space platforms. The recent Mars Pathfinder mission is a well-known achievement under this new paradigm. Before Pathfinder and its Sojourner rover arrived there had been only three missions that flew by or around the planet and two very expensive Viking landers. New mission took an entirely different, cheaper and more "aggressive" approach of throwing a lander on the planet with only several airbags to absorb the impact. The Sojourner rover has only a single CPU used in its operation - an 80C85 with a 2MHz clock rated at 100KIPS. The software on the rover was written in C and does not have explicit kinematic, dynamic or functional models. However, it does contain tables, constants and equations which are derived from experience in performance with the vehicle and embody predictions for use in power and thermal management.

NASA/JPL intend to continue to use small robotic rovers for planetary surface exploration. This task is designed to expand the current science enabling capabilities of these microrovers, while increasing their range of operation. This work includes exploration of new or improved methods of mobility, manipulation, sensing, computation, and control.

Remote Agent

It's one small step in the history of space flight. But it was one giant leap for computer-kind says the home page on the Remote Agent site. For the first time in history of space flights, a state of the art artificial intelligence system is given primary command of a spacecraft. Known as Remote Agent, the software operated NASA's Deep Space 1 spacecraft and its futuristic ion engine during two experiments that started on Monday, May 17, 1999. For two days Remote Agent ran on the on-board computer of Deep Space 1, more than 60,000,000 miles (96,500,000 kilometers) from Earth.

Building upon lessons learned from missions like Mars Pathfinder, Remote Agent is intended to cut down on the large time lapses during the design and transmission of instructions from the ground crew to a spacecraft, while taking over some of the more grueling aspects of unmanned missions, particularly the painstaking monitoring, analysis, and planning often performed by the ground crews. Remote Agent is made up of three components which each play a significant, integral role in controlling the spacecraft. Planner and Scheduler (PS) produces flexible plans, specifying the basic activities that must take place in order to accomplish the mission goals. Smart Executive (EXEC) carries out the planned activities, while Mode Identification and Recovery (MIR) monitors the health of the spacecraft and attempts to correct any problems that occur. These three parts work together and communicate with each other to make sure the spacecraft accomplishes the goals of the mission.

The testing of the Remote Agent accomplished 100% of the planned objectives. In the experiments, it operated selected subsystems based on plans formulated on board. Injection of four simulated faults tested it�s ability to resolve or work around different classes of problems, and in each case it devised the correct response. A bug in the executive interrupted the first experiment, but the bug proved to be easily correctable for future uses of the technology. For the freshest information, see the Mission Log - Voyage of Deep Space 1.

Failures

Things aren't going as planned all the time even at NASA. A failure to recognize and correct an error in a transfer of information between the Mars Climate Orbiter spacecraft team in Colorado and the mission navigation team in California led to the loss of the spacecraft. The peer review preliminary findings indicate that one team used English units (e.g., inches, feet and pounds) while the other used metric units for a key spacecraft operation. This information was critical to the maneuvers required to place the spacecraft in the proper Mars orbit. Let's hope that some future version of the Remote Agent or similar software will prevent such mistakes...

Near Future: Deep Space Missions

The Mars Microprobe Mission, also known as Deep Space 2 (DS2), is the second deep-space technology-validation mission in NASA's New Millennium Program. The mission will test key technologies for 21st-century missions, in which multiple landers released from a single spacecraft will carry out comprehensive observations of dynamic, complex phenomena such as climate systems and seismic activity. The New Millennium microprobe technologies will enable a wide range of scientific studies that would not be affordable using conventional technologies. On its arrival at Mars, in December 1999, the Mars Surveyor Lander detaches from the spacecraft and heads toward a soft landing on the surface to complete its mission. The microprobes will crash onto the Martian surface at a velocity of about 200 meters per second. Shattering on impact, each aeroshell releases a miniature two-piece science probe that punches into the soil to a depth of up to 2 meters. The microprobes' primary science goal is to determine if water ice is present in the Martian subsurface - an important clue in the puzzle of whether life exists, or ever existed, on Mars. The tiny science stations will also measure soil temperature and monitor local Martian weather. The microprobes will operate for just about 50 hours, but their designers are looking much further ahead - years into the next century.

Robotics

Scientists at NASA's Ames Research Center, Moffett Field, CA, are developing an autonomous robot to support future space missions. About the size of a softball, the Personal Satellite Assistant (PSA) will be equipped with a variety of sensors to monitor environmental conditions in a spacecraft such as the amount of oxygen, carbon dioxide and other gases in the air, the amount of bacterial growth, air temperature and air pressure. The robot will also have a camera for video conferencing, navigation sensors, wireless network connections, and even its own propulsion components enabling it to operate autonomously throughout the spacecraft. The little round robot's compact design will enable it to operate in the cramped confines of the Space Shuttle's flight deck and Space Station modules, while keeping out of the astronauts' way. Since it will operate autonomously, the astronauts' hands will be free for other tasks. For more information on robots at NASA Ames Research Center visit their home site at http://ic.arc.nasa.gov/ic/AutonomyRobotics.html.

More (or Less) Distant Future

Upcoming missions will require even higher degree of autonomy and intelligent behaviour. The goal of Deep Space 3 is to demonstrate technology needed for future interferometer missions. Some of them will search for Earth-class planets around nearby starts, resolving continental masses on their surfaces. Deep Space 4 will land on, and return a sample from a comet, having to resolve problems such as the extreme unpredictability of the cometary environment. Pluto-Kuiper Express is a robotic reconnaissance mission to explore the most enigmatic planet in our solar system. It will employ a new concept known as Beacon Operations, along with high level of autonomy for fault protection and detecting important scientific events. However, the title of the mission with the most remarkable set of goals goes to the proposed Europa cryobot/hydrobot testing. That Jupiter's moon and Mars are considered the most likely sites for extraterrestrial life. This mission involves landing a device on Europa's surface, penetrating the ice (which may be several kilometers thick) with a 4.5-foot-long, 6-inch-diameter heated probe that would carry a small tethered submersible vessel that could radio back a chemical analysis of what it found. The plan is to test the cryobot first at Lake Vostok, a freshwater lake beneath the ice of Antarctica.

As for the other AI-related projects at NASA, Dr. David Noever at NASA's Marshall Space Flight Center plans to develop "Book of Life" technology to identify and classify the tiniest life forms found on Earth and in samples from Mars. The project recently started under a grant from NASA's Advanced Concepts Office in Washington. The basic idea, as described by Professor D'Arcy Thompson of the University of St. Andrews in Scotland and his 1917 book On Growth and Form, applied the concepts of mathematics to the differences of form observed in various living things. He classified organisms by their geometric shape and found that changes of shape between species could be visualized by altering mathematical functions. It is now one of at least four principal criteria in analyzing the origin of astrobiological samples.
Another interesting concept discussed recently by Dr. Noever and his colleague, Subbiah Baskaran, describes applying natural selection to computer spacecraft design.