Fast, Cheap and Out of Control: A Robot Invasion of The Solar System
Fast, Cheap and Out of Control: A Robot Invasion of The Solar System
INTRODUCTION
Over the last four and a half years the MIT Mobile Robot
Group has pursued the goal of building totally
autonomous mobile robots for a variety of tasks. We
have refined hardware and software tools so that we can
quickly build robust interesting robots. For instance
Genghis, a six legged walking robot shown in Fig. 1
was completed 12 weeks after initial conception, in
response to a JPL workshop on micro spacecraft [1]. The
robot [2,3] was principally built and debugged by two
people, with occasional supporting help from about half
a dozen others. The robot weighs less than a kilogram
and can scramble over very rough terrain. A follow-on
vehicle [4] will be able to climb metre high rocks, and
travel at around three kilometres per hour. Such easy to
build high performance robots suggest some new ways
of thinking about planetary exploration.
Two of the principal costs in planetary surface
exploration missions arise from the mass of the
planetary rover upon launch, and hand construction of
the unique vehicle itself. In this paper, we demonstrate
that technology has progressed to the stage where we
can tackle both of these problems simultaneously b y
creating swarms of totally autonomous microrovers i n
the I to 2 Kg range. This way, total mass delivered to the
planetary surface is minimised and in addition, the
multiple copies of the rovers increase the chance of the
mission's success. Cost savings in terms of construction
dollar per Kg result, due to the opportunity to apply
mass production techniques to the roved manufacture.
CREATING INTELLIGENCE
3.
Allen
The first layer let the robot avoid both static and
dynamic obstacles; Allen would happily sit in the middle
of a room until approached, then scurry away, avoiding
collisions as it went. The internal representation used
was that every sonar return represented a repulsive force
with an inverse square drop off in strength. The vector
sum of the repulsive forces, suitably thresholded, told
the robot in which direction it should move. An
additional reflex halted the robot whenever there was
something right in front of the robot and it was moving
forward (rather than turning in place).
The third layer made the robot look (with its sonars) for
distant places and try to head towards them. This layer
monitored progress through odometry, generating a
desired heading which suppressed the direction desired
by the wander layer. The desired heading was then fed
into a vector addition with the instinctive obstacle
avoidance layer. The physical robot did not therefore
remain true to the desires of the upper layer. The upper
layer had to watch what happened in the world, through
odometry, in order to understand what was really
happening in the lower control layers, and send down
correction signals.
3.3
Herbert
Seymour
Genghis
3.6
Squirt
4.
4.4
5.1
New Vistas
CONCLUSION
REFERENCES
1 . R.M. Jones, Micro Spacecraft for Space Science, Workshop,
JPL, July 1988.
2 . R.A. Brooks, A Robot that Walks: Emergent Behaviour form a
Carefully Evolved Network, Neural Computation , 1:2, Summer
1989.
3 . C.M. Angle, Genghis, a Six Legged Autonomous Walking
Robot, MIT S.B. Thesis in Electrical Engineering and Computer
Science, March 1989.
4 . C.M. Angle, Attila's Crippled Brother Marvin, MIT Term
Paper in Electrical Engineering and Computer Science, May
1989.
5 . R.A. Brooks, A Robust Layered Control System for a Mobile
Robot, IEEE Journal of Robotics and Automation,
RA-2,14-23, April 1986.
6 . J.H. Connell, A Colony Architecture for an Artificial Creature
MIT PhD. Thesis in Electrical Engineering and Computer
Science, June 1989.
7 . R.A. Brooks and J.H. Connell, Asynchronous Distributed
Control System for a Mobile Robot, SPIE Vol. 727, Mobile
Robots, Cambridge, MA, 77-84, November 1986.
8 . J.H. Connell, Creature Building with the Subsumption
Architecture, IJCAI-97, Milan, 1124-1126, August 1987.
9 . R.A. Brooks, J.H. Connell and P. Ning, Herbert: a Second
Generation Mobile Robot, MIT Al Memo 1016, January 1988.
10. J.H. Connell, A Behaviour-Based Arm Controller,
Memo 1025, June 1988.
MIT A l
19.