
Unmanned Systems Roadmap 2007-2032
Chapter 6 Technologies for Unmanned Systems
Page 51
decision making processes in real time. With the introduction of unmanned systems to the force,
the definition of team member has been expanded to include unmanned systems. There are two
components to this effort: teaming between the unmanned systems and teaming between the
human and the unmanned systems. Human-robot teams provide a unique challenge, that is, how
to develop unmanned systems technologies to enable the human to predict, collaborate, and
develop trust with the unmanned system. Additional considerations are the coordination
between mounted and dismounted soldiers with respect to the exchange and hand-off of
information from, and control over, unmanned assets from one operator to the other.
6.6.7. Data Interfaces
Information exchanges occur primarily between the unmanned system, its control station, and
specially designed external interfaces, such as Air Traffic Control (ATC) and video feeds.
Unmanned system products, after being processed, flow to external nodes from the control
station servers through network connections. In its current form, the CDL communications
system provides a closed circuit between the unmanned system and its control station to carry
commands, status, and sensor products. As an edge device on the GIG, the control station then
provides this information to the user community while keeping the unmanned system isolated
from the GIG. CDL-equipped unmanned systems must transition from a closed circuit (or
merely using communications services) to a network node (or actually providing
communications services).
The first step to achieving network-centricity involves network-enabling the interfaces. In other
words, Internet Protocol-based network connections and routers between unmanned subsystems
and the on-board data link must be created with corresponding network interfaces between the
control station data link, control station subsystems, and the GIG. This changes the paradigm
from a closed circuit to a network node. Functions and products of unmanned systems
implemented as network nodes would be accessible to other authorized nodes on the GIG, not
just to the control station. The unmanned system itself becomes an edge device on the GIG.
The second step involves unmanned systems that can connect directly to more than one node on
the GIG. During times when the demand on the data links is low, such as during cruise portions
of the mission, unmanned systems capable of connecting to more than one node can act as
network routers, passing Internet data packets between the multiple connected nodes. In this
way, unmanned systems can contribute their unused bandwidth to the overall carrying capacity
of the GIG.
6.6.8. Dynamic Obstacle/Interference/Collision Avoidance (Including Humans)
All unmanned systems except the smallest special purpose vehicles must have the ability to
autonomously avoid obstacles. In addition to the simple avoidance of obstacles (which is not
simple if both the “obstacle” and the vehicle are moving independently), we must consider
perception elements impacting trafficability, tactical maneuver, and mission execution. While
most control algorithms are sufficiently mature, sensor processing is lacking for autonomous
operations. Some combination of radar, optical, and infrared (IR) sensors will likely be required;
and image processing algorithms, especially for the latter two, are in their infancy. Most of the
mission capabilities also require the autonomous avoidance of threat systems, including ships,
boats, craft, active sensor systems, and, to whatever extent possible, passive detection systems.
The community would benefit greatly from increased developments in this area.