Proceedings Article | 3 May 2007
KEYWORDS: Sensors, Reconnaissance, Microelectromechanical systems, Algorithm development, Robotics, Video, Sensor fusion, Safety, Satellites, Feature extraction
Robotic reconnaissance operations are called for in extreme environments, not only those such as space, including
planetary atmospheres, surfaces, and subsurfaces, but also in potentially hazardous or inaccessible operational areas on
Earth, such as mine fields, battlefield environments, enemy occupied territories, terrorist infiltrated environments, or
areas that have been exposed to biochemical agents or radiation. Real time reconnaissance enables the identification and
characterization of transient events. A fundamentally new mission concept for tier-scalable reconnaissance of
operational areas, originated by Fink et al., is aimed at replacing the engineering and safety constrained mission designs
of the past. The tier-scalable paradigm integrates multi-tier (orbit atmosphere surface/subsurface) and multi-agent
(satellite UAV/blimp surface/subsurface sensing platforms) hierarchical mission architectures, introducing not only
mission redundancy and safety, but also enabling and optimizing intelligent, less constrained, and distributed
reconnaissance in real time. Given the mass, size, and power constraints faced by such a multi-platform approach, this is
an ideal application scenario for a diverse set of MEMS sensors. To support such mission architectures, a high degree of
operational autonomy is required. Essential elements of such operational autonomy are: (1) automatic mapping of an
operational area from different vantage points (including vehicle health monitoring); (2) automatic feature extraction and
target/region-of-interest identification within the mapped operational area; and (3) automatic target prioritization for
close-up examination. These requirements imply the optimal deployment of MEMS sensors and sensor platforms, sensor
fusion, and sensor interoperability.
