Numerous modules could be used to establish a sentry network in which
the sensors cooperate to verify detections and estimate the 3D locations
and velocities of moving objects. SRI has already developed the critical
stereo sensor that makes this project feasible -- the parts cost is approximately
$200, in large quantities, and this price can be reduced significantly
if special chips are fabricated. Within the first six months of the project,
we plan to construct 20 copies of an initial stereo-based SSM, implement
a development environment for it, and then distribute sets of the SSMs
and the environment to the Integrated Feasibility Demonstration (IFD) contractor
and the other FREs as a common experimental device. Throughout the project
we will explore ways to integrate additional sensor modalities (including
acoustic and radar), enhance our moving object detection and tracking techniques,
and explore strategies for forming networks of these modules to perform
the tasks more reliably and robustly. In summary, we believe that technology
is available today to build SSMs that can save lives and provide substantial
force multiplication.
Our vision is to develop SSMs that are sufficiently competent to be used by a warfigther as an extra pair of eyes to watch his backside, while he continues with his primary mission. If a warfighter is clearing a building in a MOUT scenario, she can leave one device at the entrance to warn her if someone enters the building. If she's on a trail, she set up one on a tree to warn her of someone approaching from the rear. In addition, a warfighter can use one of the modules as a periscope to look around a corner and into a tunnel without exposing herself to enemy fire. For this application, the warfighter attaches her head-mounted display to the module's high-definition video output and visually explores the hidden area by physically moving the module to scan it. The ultimate modules will have sensors that combine infrared (IR) and image-intensified data to provide an excellent night-vision capability.
We expect the deployed modules to incorporate multiple sensor modalities, such as stereo, motion, radar, and acoustic, in order to maximize their applicability and minimize their errors. For example, a combination of a radar sensor to detect movement and a stereo sensor to construct 3D descriptions of the moving object will be one of the first devices to be considered. For some applications, a combination of acoustic and visual sensors is appropriate. For other applications, inclusion of chemical and biological sensors would add a critical new dimension.
An enhanced version of the SSM will include techniques for multiple
modules to cooperate, setting up communication channels, a common coordinate
system, and a procedure for operating together to verify detections nd
estimate their 3D locations. And our vision of the ultimate SSM is a mobile
sensor package that can be used either to explore ahead of a warfighter
or to trail behind, checking for potential danger signals.