Unexploded ordnance (UXO) characterization and remediation activities using currently available technology often yield unsatisfactory results due mainly to the inability of current navigation technology to provide precise position and orientation information. This results in poor geolocation of the acquired electromagnetic imagery and consequently the inability to discriminate between UXO and non-hazardous items in the imagery. Reliable and precise geolocation technology is essential for robust detection and discrimination of UXO in various field conditions.
The objectives of this project were to develop accurate and reliable geolocation algorithms and tools based on multi-sensor integration, to significantly improve the state-of-the-art in sensor georegistration, and to support the collection of geophysical data used to characterize UXO in various environments.
The primary technical challenges were: (1) development of a tight integration concept for Global Positioning System (GPS), Inertial Navigation System (INS), Pseudolite (PL), and Terrestrial Laser Scanning (TLS) technologies; (2) development of algorithms for ambiguity resolution for the integrated GPS/PL system; (3) testing and characterization of the stochastic properties of PL signals in order to develop optimal positioning algorithms; (4) development of methodologies and algorithms that characterize and mitigate PL signal propagation errors due to tropospheric delays and multipath disturbances; (5) development of simulation tools for the design and analysis of PL positioning geometry; and (6) development of robust algorithms for TLS surface matching and integration with GPS/INS/PL technology.
Sensor integration is based on the Extended Kalman Filter (EKF). Navigation performance and design specifications were developed for a full suite of possible deployment scenarios. System testing of hardware and software components was conducted in both laboratory and field environments. The resulting performance was evaluated against the predefined system performance specifications to define applicability and limitations of the multisensory technology for use in various UXO mapping applications.
The main achievement of this research was the integration of PL and GPS signals together with INS and TLS measurements in the form of a quadruple integration system, which can deliver high accuracy multi-sensor positioning solutions and is compatible with magnetic and electromagnetic geophysical sensors. Second, for optimal data processing, the new software, named AIMS-PRO™, was designed and implemented to incorporate the concept of a novel quadruple integration of GPS, PL, INS, and TLS. Third, precise timing of all the sensory data to GPS time was designed and efficiently implemented. Finally, a demonstration prototype was designed, implemented, and tested.
The overall system performance of the quadruple GPS/INS/PL/TLS integration system was evaluated using field data collected mainly in 2008. Test results have demonstrated the ability of the system to maintain the required accuracy level (cm-accurate positioning) in difficult environments where long GPS outages (in the range from 200 to 900 seconds) can be anticipated.
The navigation technologies developed under this project will enable accurate and reliable operation of new and emerging UXO detection sensors, where precise sensor location and orientation are required. Improved performance of UXO geolocation systems will significantly aid the inversion of field data for more accurate characterization of candidates for buried munitions, thus greatly reducing the number of false positives, which currently can reach 90% or more.