by David Geer
Contact the author at geercom@windstream.net
University of Texas UAVs,
AUVs, and UGVs
Scientists at the University of Texas, San Antonio use swarms of effective,
inexpensive land, air, and sea robots that communicate with each other to
complete important tasks as a group effort. Roboticists refer to these
robots as Unmanned Air Vehicles (UAVs), Autonomous Underwater
Vehicles (AUVs), and Unmanned Ground Vehicles (UGVs).
Tasks include surveillance, search and rescue, and mine sweeping.
Robot swarms are the most viable option for these
tasks. It is too expensive for researchers to use a single
large, expensive robot to do the work. If a task damages
the robot, it is “game over” until someone retrieves the
robot (if this is even feasible), repairs or replaces it, and the
task begins again — by which time it may be too late. This
approach costs time, money, and opportunity since battles,
territory, and lives may be lost in the interim. The cost is
higher when sending rescue teams and members of the
military as this puts human life in harm’s way.
When a swarm of inexpensive robots goes on a mission
to surveil an area, search and rescue people, or repair or
free up expensive equipment, one or two robots sacrificed
in the process will not stop the project. The remaining
robots continue to communicate with each other and work
together to finish the project. Almost no time is forfeit, and
the project completes successfully with human lives, equipment, and territory intact. Swarm robots are an affordable
solution for the military, Coast Guard, and other services to
use repeatedly without going over budget. That is why so
many government and university robot research projects are
talking, thinking, and producing swarms, advancing the
science and testing the efficacy, as the Autonomous Control
Engineering Laboratory at the University of Texas is doing.
(Other examples include NASA, MIT, the USC Information
Sciences Institute, and the Ecole Polytechnique Federale
De Lausanne (EPFL), Switzerland. There are actually
many more.)
Swarms
Close-up of a lab airplane as it is being prepared for
an autonomous flight by three researchers.
The single defining difference between swarms of
robots and other robots is the fact that they cooperate
to complete tasks, according to Ted Shaneyfelt, et. al.,
“Towards Net-Centric System of Systems Robotics in
Air, Sea, and Land.” A number of hierarchical relationships or states that exist between the robots define
the swarms, as described by Shaneyfelt. There is not
only the cooperation level, but also knowledge,
coordination, and organization levels, which are
subservient to the cooperation level. At each level, the
robots are in a state, so to speak. At the first level, the
robots must be cooperative in order for researchers
to consider them members of a swarm. This is not
optional. At the second level, the robots are aware of
each other, or they are not aware of each other. Most
10 SERVO 05.2009