International Planetary Rover Efforts

Prepared by Mark Maimone
Robotics Institute
Carnegie Mellon University

Mars / Asteroids / Moon


Mobile robots are opening up new horizons in planetary exploration. Eventually, humans will explore the planets directly: but mobile robots can help us learn a great deal about planetary surfaces right now. The Lunokhod missions of the 1970's and the 1997 Sojourner mission have hinted at the possibilities, and today's researchers are demonstrating advanced capabilities that will allow future missions to have an even grander scale.

I have collected together below some of the resources available on the Internet that talk about planetary rover efforts. Please note that many of the links here go directly to web pages of the institutions doing the work; this page is mainly a summary of useful links, not an attempt to document all the efforts myself. It is certainly incomplete, so please let me know if you have other links or information that would be useful to add to these pages! This collection was inspired by a presentation that summarized the IROS'97 Planetary Rover Workshop.

Rover Technologies

One of the defining characteristics of a planetary rover is its ability to handle rough, unknown terrain. The locomotion system must be able to negotiate large rocks, loose soil, and steep slopes. Many types of locomotion systems have been proposed, e.g.: 4-wheel, 6-wheel, 8-wheel, 1-wheel, legged, balloons, and egg-shaped hoppers.
Communication of data and images back to Earth is critical. Possible solutions to the problem of communications include having a tether from the rover to a fixed lander/communications platform, using a short-range wireless link instead of the tether, communicating from the rover directly to an orbiting satellite relay station, and the holy grail of direct rover-to-Earth communication.
Science Payload
For purely scientific missions, it is necessary to tailor the rover's capabilities to the needs of the science package, so the science requirements need to be specified far in advance of the rover's design and construction.
Obstacle and Hazard Detection
Rovers need to learn about their environment as they drive. Internal sensors like gyros, Inertial Measurement Units, inclinometers, accelerometers, and limit switches can be used to determine when the robot has entered a dangerous position (e.g., when it might be starting to roll over). But potentially even more useful are sensors that can test the environment before the rover gets there. Sensors like stereo vision cameras, laser rangefinders, sonar, and radar sensors can provide advance warning about obstacles around the robot.
Position and Pose Estimation
One of the hardest problems in developing control software for rovers is in knowing where the rover is at all times. In unknown planetary environments, maps may not be available and landmarks may be hard to find. Internal sensors like wheel encoders, gyros, and Inertial Measurement Units can be used to determine position. Unfortunately they are subject to much error as the robot moves, since it is likely to slide on loose dirt or steep slopes, fall off of small rocks and so on. GPS systems could provide much better data, if only we had networks of GPS satellites around other planetary bodies. Pose information (i.e., "which way am I pointing") can be gleaned from star trackers, sun sensors, accelerometers, Inertial Measurement Units, wheel encoders, and vision-based landmark tracking.
Autonomous Operation
Autonomous operation is critical for planetary exploration because the communications delay between Earth and planets can be many minutes. With autonomous driving, a robot can explore a much greater distance because it doesn't have to wait for a person to decide a safe route. The rover would be able to see obstacles and recognize them on its own. The more autonomous control the robot has, the farther it can progress on its mission. It is not enough to merely detect obstacles, the rover must have a higher-level system that can determine the best way to get around the obstacle, while still attempting to satisfy a higher-level goal.
Power System
One of the most challenging aspects of building a planetary rover is finding an appropriate and adequate power source. Solar power is probably the best renewable resource, but has a low yield and results in more than 50% downtime guaranteed. Other options include batteries, fuel cells, RTGs, and wind power.
User Interface
Human controls for the robot are just as important to mission success as is the robot itself. Designing the user interface to match the specifications of the mission can dramatically improve the mission results. The ability to quickly assess and modify the vehicle state makes it possible to accomplish more tasks in a short time.
Data Logging
The rover and the ground station must both be capable of storing the data acquired during a mission. For some planetary operations, a rover may have a small number of communications opportunities (e.g., Mars Pathfinder had two 5-minute windows per day), so the ability to store data over time is critical. Ground stations on Earth must also have enough storage facilities to accomodate the complete mission results.
Space Qualification
After the launch cost, the greatest expense in the construction of a planetary rover is the space qualification process. Rover components must withstand the stresses of launch, landing, the conditions encountered while travelling to the destination planet (radiation, vacuum, heat, cold), and any planetary surface hazards (dust storms, crevasses).

Teleoperated Planetary Rovers: Then and Now

The following table combines robots that have been to other worlds with current ongoing research efforts. Follow the links to learn more about each robot.
Vehicle Organization
Lunokhod 1, 2 Soviet Union
Mars 3 Soviet Union
Mother/daughter ISAS
Marsokhod IKI
Marsokhod NASA Ames
Pebbles MIT
Nomad CMU
Rocky 7 NASA JPL
Sojourner NASA JPL
Nanorover NASA JPL
Italics indicate robots that have already been delivered to other worlds

Deployment and Field Test Summary

These days it is not enough to just develop technology on the bench, it must be demonstrated in the context of an active, mobile platform. Here are some of the most fully developed Field Tests of planetary rovers that have been documented.

Robot Location Total Distance Duration Average Instantaneous Speed Continuous
Median Max
Lunokhod 1 Moon 10.5 km 300 days 35 m/day 1970
Lunokhod 2 Moon 37 km 120 days 308 m/day 1973
Ames Marsokhod Hawaii 2 km 6 days 333 m/day 1995
Ames Marsokhod Arizona 0.5 km 6 days 83 m/day 1996
Ratler NREC Moonyard 43 km 30 days 1433 m/day 50 cm/s 70 cm/s Y 1996
Nomad Atacama Desert 224 km 46 days 4869 m/day 43 cm/s 43 cm/s Y 1997
Sojourner Mars 0.1 km 112 days 0.9 m/day N 1997
Rocky 7 Lavic Lake 1.1 km 14 days 78 m/day N 1997
Nomad Antarctica 30 days (1998)
Iares Antarctica (1998)

International Consortia

These days everyone wants to get into space, not just governments. Several industrial and government consortia have grown over the last view years. They have the novel and laudable goal of launching a planetary rover using (some, most, or all) funding from non-governmental sources.
Euromoon 2000 (nee LEDA)
The first phase would survey lunar surface regions (e.g. poles) using orbital and in-situ measurements, including a modest lunar rover. Components are being tested now, e.g. the wheels and tether for the IDD/MIDD at DLR in Germany.
Lunar Rover Initiative
LunaCorp and CMU are pursuing commercial sponsorship of a two rover exploration venture to the Moon. They would visit Apollo and Lunokhod landing sites, and travel 2000 km in two years.
Earthrise 2001
A consortium of Japanese companies (e.g., Nippon TV, Mitsubishi, Shimizu) plan to land a lunar rover in the year 2001.

Recent Workshops and Public Events

Planetary rovers are beginning to enter the public consciousness, thanks in large part to the success of NASA's 1997 Mars Pathfinder mission. Some of the events that have helped spur this interest in the last year are listed here.
NASA Mars Pathfinder Mission
From 4 July to October 1997, the Mars Pathfinder mission's Sojourner rover captured the public's imagination. This mission was the first to deploy an intelligent mobile robot to the surface of another world. Daily reports appeared in international news media, and web site hit count records were broken.
Disney Epcot Space Celebration
From 3-7 October 1997, several planetary rovers were on display at Epcot Center in Disneyworld, Florida, USA.
IROS'97 Planetary Rover Workshop
On 7 September 1997, the IROS Planetary Rover Workshop was held in Grenoble, France. Results from several rover research groups were presented, and a summary of the presentations is available here. For more information, please contact the workshop organizer, Dr. Richard Volpe of JPL.
1997 Atacama Desert Trek
From 15 June to 31 July 1997, visitors to the Carnegie Science Center in Pittsburgh had the opportunity to test-drive the Nomad robot through the Atacama Desert in northern Chile. Part of the CSC's Robotics exhibit, tens of thousands of visitors not only got to see live images from the Atacama, but also got to pick the viewing direction. A lucky few even got to drive the robot. Many more got to participate by viewing the project's web pages, viewing live imagery on the web and reading the daily field reports.
Lavic Lake Field Test
From 22-30 May 1997, several high schools got the opportunity to drive the Rocky 7 robot over desert terrain in California.
1997 Rover Roundup
On 1 February 1997, the Planetary Society sponsored a planetary Rover Roundup in Santa Monica, California, USA.

Related Links

  • Chinese Rover
  • Rationale for Mobility in Planetary Environments, report by the National Research Council
  • NASA Space Telerobotics Program
  • Robotics Internet Resources
  • Planetary images courtesy of Views of the Solar System web site
  • FPSpace rover query results: Mars-3, hopper
  • Apollo 15 Lunar Roving Vehicle
  • Soviet Lunar missions (Lunokhod 17 and 21)
  • Apollo series Lunar Rovers
  • Japanese Lunar Base concepts, another report
  • Japan: To the Moon and planets
  • Rover Roundup 97 photo and participants
  • LAAS Robots
  • Viacheslav Linkin, Marsokhod scientist
  • Lavic Lake Results and paper rough draft.

  • Last modified 11 November 1997