The Retelling of
“Romancing the Rover (How Sojourner Came to Be in the Late 1980s and Its Journey to
JPL Stories, September 28, 2000
Presenter: Don Bickler
Reteller: Marilyn Morgan with Don Bickler
The sixth in the Library’s series of JPL Stories attracted a roomful of enthusiasts, including some
of the presenter’s team members ―to see,‖ he asserted, ―if I make a mistake.‖ The room buzzed
with anticipation, and even before Don Bickler began his talk, we could perceive a familiar JPL
gestalt — the energetic demeanor, the ebullient manner, the general look of someone who is
about to share something that is really terrifically exciting.
Sojourner had gone to Mars, fulfilled all expectations, captured the public imagination, even
became a toy, and finally emerged as an icon of JPL ingenuity. We were about to hear from
someone who was present at the creation — and not only that, he was wielding the wrench and screwdriver.
I’m convinced that engineers, despite their sober and serious reputation, have just about all the
fun there is to be had in the world. They grab onto a problem and address, define, analyze, and
solve it intellectually. That’s exhilarating in itself, but then they get to build things: structures,
bridges, vehicles, rockets. Rocketry engineers have even more fun: they get to build things and
then fire them into the sky or (accidentally of course) blow them up. If you have seen the old
films of Robert Goddard and his crew jubilantly experimenting with liquid-fueled rockets in the
New Mexico desert, you know what I’m talking about.
Engineers who build robotic vehicles also have a bit of that maniacal gleam in the eye. They
experience the joyfulness of creating things that scurry, scoot, roll, reach, climb, scramble, pivot,
and peregrinate, while carrying out instructions on command or, even better, these neoteric
critters decide things on their own through software created by their human designers. Like
symphonies, sophisticated robots are art forms. But the robots that seem to operate so easily and
beautifully are not born so much as bred, through generations of robots, some of which never get
beyond the drawing board, and some of which are magnificent failures. Sojourner, which seems
Sojourner by Bickler: FINAL FINAL — page 1
the perfect realization of a compact robotic rover, evolved at JPL over quite a few years. Her distant lumbering ancestor is the Surface Lunar Roving Vehicle, circa 1967.
The Bogie Man
Don Bickler was introduced by Barbara Amago of the JPL Library storytelling team. We learned that Don graduated from Northwestern University in 1956 with a degree in mechanical engineering, and eventually ended up at JPL in 1975. He is now supervisor of the Advanced Mechanical Systems Group in the Mechanical Engineering Section (352). Bickler’s invention of the rocker-bogie suspension system earned him a nickname — The Bogie Man.
(An aside — According to the Oxford English Dictionary, a golfing ―bogey‖ derives from a
popular 1899 song describing a dreaded person, devil, or goblin, but a mechanical ―bogie‖ originated in 19th-century northern English dialect, describing a low truck on four small wheels used by masons to move large stones. Eventually ―bogie‖ came to describe a truck running on two or more pairs of wheels supporting the forepart of a locomotive engine on which it swivels freely in passing curves.)
A before-talk exchange between Don and an audience member confirmed that the bogie design is commonly used in tandem-wheeled trucks, military trucks, and cranes to provide stability and mobility on uneven terrain. Could Sojourner, so refined looking, be related to these rough characters?
In the 1960s, we were told, the military developed several types of articulated vehicles. Don showed photographs of these strange vehicles, including a six-wheeled ―bent jeep‖ that was
supposed to be able to climb over low obstacles, and the Lockheed Twister, bearing the appearance of several vehicles in a state of post-collision. Another, the sturdy-looking Gamma Goat, was classed as a 1-1/2-ton truck and sported articulated six-wheel drive. All these were designed to traverse rough terrain such as might be encountered in war-fighting situations or military logistical enterprises. Early rover designers studied these types of vehicles for clues to creating mobile machines that would be successful in planetary excursions.
Sojourner by Bickler: FINAL FINAL — page 2
Lunar and Planetary Roving Vehicles
Sojourner’s pedigree begins with a Surface Lunar Roving Vehicle (SLRV) delivered to JPL about
1967. The segmented vehicle was supposed to fold up into an ―S‖ shape, and could climb
obstacles 1.5 times its wheel diameter. The SLRV was never launched, but provided a study
object for early JPL rover engineers. The vehicle was invented by Mieczyslaw G. Bekker, whose
team designed the Lunar Roving Vehicle used in the Apollo 15, 16, and 17 missions (1971–1972).
Bekker’s first book on the theory of land locomotion and terrain–vehicle interactions
(―terramechanics‖) was published in 1956, and by 1963 he was proposing Moon rover designs.
Bekker tried lots of things in his designs, including variations on Archimedes’ screw.
In those days — the heady Apollo days — NASA was thinking big, and cost was not a problem. The Apollo LRV was about 10 feet long by 6 feet wide (to wheel centers). Each spring-steel,
wire-mesh wheel was driven by a 1/4-hp electric motor, and the front and rear wheels had
separate steering systems. Either of the front or rear wheels could be turned while the other pair
was locked straight, or both could be used simultaneously, so the rover could turn completely
around within its own length. It could climb over obstacles 1 foot high and negotiate 25-degree
slopes. All these were desirable characteristics in a roving vehicle. Additionally, the Apollo LRV
had to fit into the Lunar Descent Module — for stowage, all four wheels folded inward and the vehicle folded double — and be easily deployed on arrival.
Another idea for a lunar rover was the MOLAB or mobile laboratory: a 3-ton, closed-cabin
vehicle with a range of about 100 kilometers. The MOLAB was an articulated vehicle with
independent suspension and 1.5-meter-diameter wheels, but was far too large for what NASA
ultimately required for Apollo.
Big Mars Rover Concepts
In the 1980s, interest in Mars spurred new thinking at JPL about what we would want in a robotic
rover, should the opportunity arise for a journey. The SLRV, the baseline vehicle, morphed into
the Little Blue Rover. The Blue Rover had an elastic (spring) chassis and could clamber over 1-
1/2-wheel-diameter obstacles. Don showed us a vintage-1986 drawing of a rover concept that
illustrated a six-wheeled articulated vehicle of three sections: a forward portion with two science-
instrument arms, a middle portion with camera, and an aft portion that seemed to be solely RTGs
on two wheels. (The shape of the vehicle wheels resembled narrow automobile tires, and this
Sojourner by Bickler: FINAL FINAL — page 3
became an issue when different types of terrain were considered.) Bickler began thinking about
bogie linkages to overcome the undesirable characteristics of using springs.
The next illustration Don showed us was ―Fig. 3‖ from his 1989 patent (no. 4,840,394) for an
articulated, springless suspension that maintains uniform distribution of weight and traction on
the wheels. Fig. 3 showed a six-wheeled vehicle with front and rear linkages coupling the front
and rear wheels to the middle wheel. It was assumed that equal weight on each wheel is best, but
the trick was to make three linkages work (―like clapping with three hands‖). The design has a
master link on each side, but there was a problem — you can’t steer the wheels. The next drawing
was a wire frame of a six-wheeled vehicle based on a pantograph design. (A pantograph is the
drawing instrument made of four rigid bars joined in parallelogram form to maintain scale while
copying a drawing.) The linkages allow steerable struts. Now we’re getting somewhere.
In a JPL review discussing rover designs, a CAD of the pantograph design drew pointed criticism
from John Casani — ―too complicated,‖ ―we don’t know if it will work,‖ and ―why not four wheels?‖ That shout from the back of the room was Bickler’s: ―How about a model?!‖ Don’s 1/8-
scale table-top model for the pantograph was favorably received at the next review.
In the search for balance among the trade-offs, considerations of weight on the wheels, size of the
wheels, number of wheels, type of wheels, steering and articulation, surface speed, ground
clearance, and how high the vehicle could climb (size of the obstacle to overcome) were all part
of the picture. Four wheels were not enough; six was the optimum number of wheels for stability
and obstacle-crossing capability. The idea was that a rover would have to be able to climb over
obstacles 1.5 meters high — almost 5 feet. Objects of this size could be seen from orbit. Don’s photograph of a person holding a 1.5-meter-long measuring stick showed us that such a rover
would have to be really large, and indeed, the next photo showed an experimenter in the east
Mojave Desert, standing, for scale, next to wheels of the requisite size: huge, 1-meter wheels on a
single axis. These experimental wheels were narrow in shape. Wheel shape is something to think
about; you don’t want your rover to start off on a Martian traverse and get ignominiously stuck in
a ravine into which the wheels precisely fall. Don recalled riding a bicycle as a kid in Chicago
along streetcar tracks and realizing that ―if you get stuck in the tracks, you’re dead.‖ This insight,
he declared, ―really made an impression on me.‖
Sojourner by Bickler: FINAL FINAL — page 4
between the wheels. As the trailer is pulled by the tractor, the trailer’s wheels rise vertically in
attempting to negotiate the bump. But the trailer’s wheels lock up and are restricted from going
forward; at the same time, they pull the tractor backwards. The tractor wants to keep going, and
its second set of wheels have a tendency to drive the vehicle up under itself, raising the front
wheels. This was nicknamed ―popping a wheelie.‖
The problems for a rover, expected to meet plenty of bumps on Mars, were evident. If you could
maintain the wheel load distribution without resorting to the use of an oscillation-producing
spring suspension, you would increase the stability and terrain-negotiating performance of the
vehicle. Through optimizing the geometry of a bogie system, Bickler produced the (now-
legendary) rocker-bogie suspension design. No one is exactly sure who named this bit of
engineering history: a six-wheeled chassis with a springless, articulated system that maintained
even weight and traction on all the wheels — the simplest mechanism to satisfy the requirements
without pantograph linkages
The Iterations of Rocky Rover
Don showed us photographs of the iterations of Rocky, as the robotics team experimented with
electronics, batteries, materials, and loads. They even went to a farm equipment show to learn
about struts. Rocky I, with 9.25-centimeter (3.66-inch) diameter wheels, could traverse steps and
gaps over two wheel diameters, and climb over rocks about twice its size. The Rocky II design
was abandoned before realized in hardware, but Rocky III was another improvement on the
rocker-bogie design, with more mobility and lower surface pressures. Rocky III had 13-
centimeter (5-inch) wheels, the same dimensions as the 1/8-scale table model of yore. (Bigger
wheels work better in sand; smaller wheels sink.) Don showed us a photograph of a ―dramatically
overloaded‖ Rocky III being tested in the Mojave Desert, and here we observed the rover
surrounded by robotics team members bearing the bemused expressions one sometimes sees on
parents’ faces as they watch a tottering but determined offspring engage the uneven terrain of the
world. The vehicle, with all its apparatus, was so heavy that the metal wheels turned in the rubber
tires — the experimenters had to use Crazy Glue to attach the tires to the wheels.
Sojourner by Bickler: FINAL FINAL — page 6
The slimmed-down, sophisticated Rocky IV was a flight prototype version of Rocky III, with all-
metal (rubber is not a desirable material on the cold surface of Mars) flexible wheels to get
around better in sand and other soft materials. But the slotted wheels had a tendency to break and
protrude, locking against the struts, so the next wheel design was solid metal. An iteration or two
more, including Rocky VII, a close-coupled design intended to have one motor drive two wheels,
and finally Rocky VIII — the elegant little rover that became Sojourner.
The naming of the rover, Don explained, was Donna Shirley’s idea, and ―it had to be female.‖ (Donna Shirley was manager of the JPL Mars Exploration Program.) When informed that the
rover would be named thusly, Don impishly proposed the name ―Trixie.‖ Donna affixed him with
a look (and one can only imagine what that gaze conveyed), and proclaimed, ―You’ve got nothing
to do with it, Bickler!‖ A playful name was not to be; instead, the rover was eventually named in
honor of the remarkable 19th-century American abolitionist and champion for women’s rights,
During the question and answer portion of the storytelling, someone asked about optimum wheel
size: how does one get there? It turns out that the wheel size is selected through exploring the
solution space using a blend of human judgment and computer analyses. As Don pointed out, ―a computer programmed to overlook flotation in soft sand would say that zero-diameter wheels are
best.‖ As to whether a Sojourner could be created in the ―faster-better-cheaper‖ environment,
Don’s answer was ―Absolutely!‖ It’s engineering interest, not the prosaic limits of 40-hour weeks,
that drives the process, and ―we have a very big LEGO kit‖ for experimentation.
Another questioner asked about the biggest threat to the rover, and that, as might be surmised, is a
situation where the rover somersaults. The rover allows for more than a wheel diameter of
obstacle traversal, and can be symmetrical in forward/backward motion capability. A hazard-
detection system is designed to prevent an overtipping condition by sensing pitch and roll; if pre-
set limits are exceeded, the rover stops. Sojourner can actually climb a 45-degree slope (if the soil
can be piled that steep), and can scale a 20-centimeter-high rock. If Sojourner were somehow
flipped over, though, it could not right itself. Information on traversable obstacles (and the
allowed traversability numbers are conservative) is stored on board, and if Sojourner encounters a
non-traversable object (a rock too high), it is supposed to turn away and find another path. An
Sojourner by Bickler: FINAL FINAL — page 7
interesting example of Sojourner meeting and evaluating an obstacle on Mars is captured in a
brief movie at <http://mars.jpl.nasa.gov/MPF/ops/rvrmovie.html>. Open up the
first item on the list, <rover_movie_sol24_S0050Q.gif>, and you’ll see the rover bump a
pyramid-shaped rock, turn, rotate 35 degrees, back up, and then blithely move forward as its right
set of wheels clambers right over the rock. (Sojourner is capable of climbing over rocks twice that
size.) The movie is a neat little example of the rover’s climbing abilities, and provides a good
view of how the suspension system and steering work.
Then there was the sand issue — could Sojourner get stuck? Don gleefully told the story. The
team had tested the rover in sand created to ―match Houston’s recipe‖ for lunar soil, so they knew
the rover was capable of traversing the fluffy stuff. A (nameless) challenger, citing M.G.
Bekker’s equations, maintained that the JPL rover was such a poor design that it could not handle a downhill slope. The challenger was politely invited to come to JPL, where a hole was dug in the
simulant for the rover. Sojourner, of course, promptly climbed right out.
At the end of the too-short hour, Barbara Amago thanked Don and presented him with a small
crystal globe, the customary gift that represents the jewels of insight brought to us by the JPL
Eyeing it appreciatively, The Bogie Man pronounced it ―one heck of a marble.‖
Web Sites of Interest
Timeline of JPL Mars rover development.
http://mars.jpl.nasa.gov/MPF/rovercom/pixt.html — Mars Microrover photo
gallery. Here are many pictures of various rovers as well as photos from the Pathfinder mission.
Includes a composite image comparing the size of Sojourner with the Apollo Lunar Roving
http://www.delphion.com — The 1989 rocker-bogie patent. Type in ―Bickler‖ and a list of patents comes up, including no. 4,840,394 for the 1989 articulated suspension system.
Sojourner by Bickler: FINAL FINAL — page 8
http://mpfwww.jpl.nasa.gov/rovercom/rovintro.html — Information on Sojourner (Mars Microrover).
http://www.bchip.com/mars/rover/descrip.html — Detailed description of Sojourner rover.
http://mars.jpl.nasa.gov/MPF/ops/rvrmovie.html — ―Movies‖ of Sojourner on Mars.
http://www.bchip.com/mars/rover/name.html — How Sojourner rover got its name. http://www.sojournertruth.org/ — The Sojourner Truth Institute Web site.
http://www.digitalsojourn.org/speech.html — Sojourner Truth’s famous 1851 speech.
http://www.osa.com.au/~cjh/lego/ — An Australian hobbyist’s palm-sized Sojourner model built of LEGO blocks, with a miniature rocker-bogie system.
Sojourner by Bickler: FINAL FINAL — page 9