Environmental Engineering Reference
In-Depth Information
evolved into Distributed Interactive Simulation (DIS), and then into High
Level Architecture (HLA) [
18
].
These systems created a shared virtual environment where the combat
elements (tank, plane, missile, helicopter, etc.) can see themselves and the
other combatants. Each soldier sits at a station that controls an element (e.g.,
a tank position or the cockpit of a fighter) and the soldier's actions cause an
appropriate change in the simulation of the element in the virtual world. The
soldiers are given a view appropriate to their vehicles and stations and they
are able to see the other combatants and the effects of their actions (a missile
being fired or a tank turret being rotated). The system has been deemed so
good that it has been used to test out new tactics and has been used to assist
in the design of new systems by allowing different designs or tactics to be
simulated and tested under simulated combat conditions.
Hundreds of individual vehicle simulations can be connected together over
a distributed network using specialized protocols running over Internet proto-
cols. These protocols support the ecient exchange of simulation information
and allow all participants to experience an appropriate view of the virtual
world without requiring an overwhelming amount of computation per station
or overloading the network with simulation updates. Work has been directed
toward building simulated forces, linking real physical hardware directly to
simulated hardware, and building a virtual environment that would allow
foot soldiers to engage in simulated combat.
DIS and HLA have been highlighted because they represent the high end
of software testing environments for cooperative autonomy. They can support
large numbers of simulated objects in a physically distributed environment
using Internet protocols. They can also support the integration of real hard-
ware with simulations. If the HLA protocols were modified to meet NASA
requirements, the resulting system could allow detailed testing of proposed
cooperative autonomy systems, or could allow realistic ground support sta-
tions to be integrated into the environment to test new control regimes or to
train ground support personnel.
7.6.5 IBM Aglets
Many different technologies have been proposed to support agent-based pro-
gramming. One system developed by IBM supports the creation of Aglets,
which extends Java-based applets to create mobile software agents [
82
]. The
Aglet toolkit helps the programmer develop autonomous agents, which can
then be instantiated, cloned, moved to other computation systems, or de-
stroyed. Implemented in Java, the Aglets have an advantage that they can
run on any computation platform that supports Java, and they automatically
have the many security features provided by Java. The Aglet toolkit does not
focus on cooperation between Aglets, but these services could be provided
by other Java classes. IBM has made the Aglet toolkit publicly available to
support experimentation by others.
