This PDF file contains the front matter associated with SPIE Proceedings Volume 7333, including the Title Page, Copyright information, Table of Contents, Introduction (if any), and the Conference Committee listing.

Even with traditional system design and development programs military systems are developed that end up being difficult for the target audience to use. Over the years the military has learned to incorporate human factors considerations and requirements in system requirements documents in order to minimize this problem. However in today's environment of procuring GOTS/COTS equipment to quickly field a needed capability, the human factors aspects are not always considered or they may have to be traded for other considerations. This occurs for a variety of reasons with the driving reason being the willingness of commanders and agencies to trade capabilities for speed of fielding. This paper addresses human factors considerations that should be observed in the design of unattended ground sensors (UGS) at the component, equipment and system levels. This is not an abstract paper on human factors engineering but an examination of current trends and applications. Lessons learned from recent fieldings and example designs from the Harris Falcon Watch system are provided. What Harris has found is that design considerations, development schedules, understanding of the target audience and the mission scenarios, and training are all key factors in determining whether a system will be found to have utility by a broad spectrum of users.

McQ has developed and delivered numerous unattended ground sensor (UGS) systems for a variety of applications. The systems provide flexible, wireless communications and numerous options for enabling the user to configure the system for a specific mission. This flexibility is a two-edged sword as it provides both the intended user with the functionality they desire, but also a set of vulnerabilities if a malicious user (e.g. political enemy or competitor) would attempt to disable or reverse engineer the system. McQ has developed various layers of security to address: secure program and data storage on off-chip non-volatile memory; secure access to JTAG on COTS processors and DSPs typically incorporated in the design of embedded systems used for remote sensors; authentication of sensors nodes, relays, and portable user interfaces used in the field that may be compromised; and the management of keys and other security-related data that is required to be stored and maintained in a distributed system. The associated challenges with securing embedded systems typically found in UGS will be described, as well as an overview of the solution that was developed and incorporated into McQ's systems to mitigate the vulnerabilities.

This paper addresses improvements and benefits derived from the next generation Northrop Grumman SCORPION II family of persistent surveillance and target recognition systems produced by the Xetron campus in Cincinnati, Ohio. SCORPION II reduces the size, weight, and cost of all SCORPION components in a flexible, field programmable system that is easier to conceal, backward compatible, and enables integration of over forty Unattended Ground Sensor (UGS) and camera types from a variety of manufacturers, with a modular approach to supporting multiple Line of Sight (LOS) and Beyond Line of Sight (BLOS) communications interfaces. Since 1998 Northrop Grumman has been integrating best in class sensors with its proven universal modular Gateway to provide encrypted data exfiltration to Common Operational Picture (COP) systems and remote sensor command and control. In addition to being fed to COP systems, SCORPION and SCORPION II data can be directly processed using a common sensor status graphical user interface (GUI) that allows for viewing and analysis of images and sensor data from up to seven hundred SCORPION system Gateways on single or multiple displays. This GUI enables a large amount of sensor data and imagery to be used for actionable intelligence as well as remote sensor command and control by a minimum number of analysts.

Unmanned Ground Sensors (UGS) have seen resurgence in recent years for use in a growing number of remote surveillance applications. These sensors can provide a wide range of information to assist an analyst in recognizing the type of intrusion detected. The addition of sensor cued imagers has also gained popularity in extending the recognition capabilities of sensors to allow identification of people and vehicles thereby expanding the mission capabilities of these systems. We are now on the brink of the next advance in remote surveillance - unmanned video - which promises to provide information far beyond recognition and identification of individuals. Knowledge of the intent of individuals operating within an Area of Interest (AOI) is possible with the retrieval of collected video. Three technologies are converging to drive remote video capability; (1) low power video processors allow advanced video functions including video compression and automated target tracking to be applied at the video input point, (2) high bandwidth tactical radio networks offering robustness and communication range beyond commercial networks are now available to exfiltrate the video, and (3) low power sensors provide the ability to maximize system operational life through power management of multiple tiers within the system. These advances have combined to create the Remote Video Surveillance Systems which promise a leap forward in the situational knowledge provided by unmanned systems.

Ultra Electronics has developed a gunshot location sensor small enough to be rifle-mounted. It measures range and direction (bearing and elevation), providing an immediate benefit to the soldier who is carrying it. It is small enough to be fitted unobtrusively on a wide range of other mobile or fixed platforms. The sensor operates standalone, but can also be wirelessly networked, potentially enabling real-time plotting of live-fire contacts, which could revolutionize the commander's view of an engagement and his ability to concentrate firepower. This paper describes the technology, the results of trials, and the potential applications for such a sensor.

The Army currently employs heterogeneous unattended ground sensors (UGSs) using a sparse deployment to maximize coverage, minimize pilferage and to monitor terrain bottlenecks. A team consisting of Teledyne Scientific Company, the University of California at Santa Barbara and the US Army Research Laboratory (ARL) is developing technologies in support of automated data exfiltration from heterogeneous battlefield sensor networks as part of a US Army contract¹ with the Institute for Collaborative Biotechnologies (ICB). The ICB program is developing a new system consisting of novel bio-inspired software algorithms for autonomous operations that will leverage proven research to monitor sensor networks from extended ranges, that will collect data in a timely fashion, that will collaboratively control the motion of a sparse network of collectors (e.g., UAVs) using bio-inspired sampling, that will accurately detect and localize field events and will fuse and classify sensed data. A new bio-inspired event discovery technique will enable fusion of sensor observations at low SNR without requiring a prior model for the event signature; this is a first step towards sensor networks that are capable of learning. The program will also provide both laboratory and field demonstrations of these capabilities supported through ARL by leveraging available resources.

The Joint Force Protection Advanced Security System (JPFASS) is a Department of Defense effort to improve conventional force protection. It is sponsored and managed by Joint Program Manager - Guardian (JPM-G). The main objective of JFPASS is to provide an integrated and layered base defense system, which includes data fusion, Command and Control (C2) nodes, Common Operation Picture (COP) nodes, and full integration of a selected range of robots, sensors, cameras, weapons, tracking systems, and other C2 systems. The URIM is the main integration tool for several sensors, cameras, and weapons in JFPASS. The Universal Resource Interface Module (URIM) is an extremely flexible framework for rapidly integrating new sensors into the JFPASS. Each sensor system has its own proprietary protocol, which makes integration high cost and risk. The URIM communicates directly with each sensor system though a protocol module and maintains a generic data object representation for each sensor. The URIM then performs a translation of the data into a single protocol, in this case Systems Engineering and Integration Working Group (SEIWG) ICD-0100. With this common protocol the data can be provided to a data server for publishing. Also, this allows for network control and management of all sensor systems via any C2 node connected to the data server.

Intelligence on abnormal and suspicious behaviour along roads in operational domains is extremely valuable for countering the IED (Improvised Explosive Device) threat. Local sensor networks at strategic spots can gather data for continuous monitoring of daily vehicle activity. Unattended intelligent ground sensor networks use simple sensing nodes, e.g. seismic, magnetic, radar, or acoustic, or combinations of these in one housing. The nodes deliver rudimentary data at any time to be processed with software that filters out the required information. At TNO (Netherlands Organisation for Applied Scientific Research) research has started on how to equip a sensor network with data analysis software to determine whether behaviour is suspicious or not. Furthermore, the nodes should be expendable, if necessary, and be small in size such that they are hard to detect by adversaries. The network should be self-configuring and self-sustaining and should be reliable, efficient, and effective during operational tasks - especially route surveillance - as well as robust in time and space. If data from these networks are combined with data from other remote sensing devices (e.g. UAVs (Unmanned Aerial Vehicles)/aerostats), an even more accurate assessment of the tactical situation is possible. This paper shall focus on the concepts of operation towards a working intelligent route surveillance (IRS) research demonstrator network for monitoring suspicious behaviour in IED sensitive domains.

This paper describes the NATO Task Group SET-093/RTG53/MSE (referred to as TG-53 in this report) Acoustic Detection of Weapons Firing Joint Field Experiment II conducted at the Etablissement Technique de Bourges (ETBS), Bourges, France, during 16 to 27 June 2008. This field experiment is a follow-on to the NATO TG-53 Acoustic Detection of Weapons Firing Joint Field Experiment I conducted at the Yuma Proving Grounds (YPG), Yuma, Arizona, USA, during 31 October to 4 November 2005 [1]. The objectives of the joint experiment were: (i) to collect acoustic signatures of direct and indirect firings from weapons' such as small arms, mortars, artillery, rockets, and C4 explosives, (ii) to analyze the propagation effects of grassy, wooded, and urban terrains, (iii) to share signatures collected from a variety of acoustic sensors, on the ground and in the air, distributed over a wide area, and (iv) to demonstrate the interoperability of disparate sensors developed by the various nations involved. The participating NATO countries , including France, the Netherlands, the United Kingdom, Canada, and the United States of America, and Israel as well as part of the Mediterranean dialogue countries, deployed nearly 90 sensors and sensor systems over the test range area.

As part of the NATO SET-093 experiment, Defence R&D Canada - Valcartier collected acoustic signatures using two Ferret systems. The new set of data was used to assess the performance of Ferret not only for the detection of small arms fire but also to determine whether weapons other than small arms could trigger the system and create false alarms. Ferret is an acoustic signal processing system that detects, recognizes and localizes the source and direction of small arms fire. New detection algorithms have been developed at DRDC Valcartier and incorporated into a recent software upgrade of the system. This paper presents an overview of the improvements, the reasons behind those changes and the performance of Ferret when exposed to the new set of data. The author also proposes metrics for future data collection that would allow a better evaluation of performance.

The US Army Research Laboratory has conducted experiments using acoustic sensor arrays suspended below tethered aerostats to detect and localize transient signals from mortars, artillery, and small arms fire. The airborne acoustic sensor array calculates an azimuth and elevation to the originating transient, and immediately cues a collocated imager to capture the remaining activity at the site of the acoustic transient. This single array's vector solution defines a groundintersect region or grid coordinate for threat reporting. Unattended ground sensor (UGS) systems can augment aerostat arrays by providing additional solution vectors from several ground-based acoustic arrays to perform a 3D triangulation on a source location. The aerostat array's advantage over ground systems is that it is not as affected by diffraction and reflection from man-made structures, trees, or terrain, and has direct line-of-sight to most events.

The detection and localization of artillery guns on the battlefield is envisaged by means of acoustic and seismic waves. The main objective of this work is to examine the different frequency ranges usable for the detection of small arms, mortars, and artillery guns on the same hardware platform. The main stages of this study have consisted of: data acquisition of the acoustic signals of the different weapons used, signal processing and evaluation of the localization performance for various types of individual arrays, and modeling of the wave propagation in the atmosphere. The study of the propagation effects on the signatures of these weapons is done by comparing the acoustic signals measured during various days, at ground level and at the altitude of our aerostat (typically 200 m). Numerical modeling has also been performed to reinforce the interpretation of the experimental results.

Detecting and localizing impulsive acoustic sources in the daytime using distributed elevated acoustic sensors with large baseline separations has distinct advantages over small ground-based arrays. There are generally two reasons for this: first, during the daytime, because of more direct and less encumbered propagation paths, signal levels are generally larger at altitude than near the ground. Second, larger baselines provide improved localization accuracy. Results are reported from a distributed array of acoustic sensors deployed during an experiment near Bourges, France during June of 2008. The distributed array consisted of microphones and GPS receivers attached to the tether lines of three widely separated aerostats. The sound sources were various impulsive devices. Results from the measurements are presented and discussed. Localization errors (GPS accuracy, propagation calculation, and aerostat motion, etc) are discussed. Possible ways to improve the localization accuracy are suggested.

Herein is described the U.S. Army RDECOM-ARDEC's purpose and series of activities conducted at the 2008 NATO SET-093 TG-53 experiment and field test. The overall purpose of the field test as stated by SET-093 panel was to provide a baseline test capable of providing relevant scenarios and data regarding a variety of impulsive generated acoustic events. As organized, the field experiment also allowed the room o study sensor interoperability across multiple platforms and multi-national users via the spider communication framework/reporting structure. This multinational network maintained by the host ETBS with a standardized messaging format with specific goals for each participating organization. ARDEC's role and purpose for the test was to provide situational awareness via the Spider and associated messaging format to the ETBS command center while continuing to gather unique acoustic data from various vantage points. ARDEC had several deliverables for the TG-53 field experiment derived from the mission and spirit of the field test. The most relevant deliverable was to demonstrate sensor interoperability via the Spider network and provide situational awareness by describing the said mortar/artillery events. The second purpose revolved around a relevant environment algorithm validation of the muzzle blast discrimination for future UGS transition in particular the UTAMS II. The algorithm validation information remained internal to the specific data acquisition system and not broadcasted out on the Spider network. The TG-53 field experiments provided the added opportunity to further test and refine the algorithm based on the discrete wavelet transform (DWT) and multiresolution analysis. These techniques are used to classify and reliably discriminates between launch and impact artillery and/or mortar events via acoustic signals produced during detonation. Distinct characteristics are found within the acoustic signatures since impact events emphasize concussive and shrapnel effects, while launch events are similar to explosions, designed to expel and propel an artillery round from a gun. The ensuing signatures are readily characterized by variations in the corresponding peak pressure and rise time of the waveform, differences in the ratio of positive pressure amplitude to the negative amplitude, variations in the prominent frequencies associated with the blast events and variations in the overall duration of the resulting waveform. Unique attributes can also be identified that depend upon the properties of the gun tube, projectile speed at the muzzle, and the explosive/concussive properties associated with the events. The event allows the examination of particular extreme battlefield acoustic challenges not normally documented or readily studied. The final portion will focus on the unique acoustic signatures data collected and how it allowed very relevant situations to be tested in a variety of scenarios.

An aerostat based acoustic array data collection system was deployed at the NATO TG-53 "Acoustic Detection of Weapon Firing" Joint Field Experiment conducted in Bourges, France during the final two weeks of June 2008. A variety of impulsive sources including mortar, artillery, gunfire, RPG, and explosive devices were fired during the test. Results from the aerostat acoustic array will be presented against the entire range of sources.

This paper presents an algorithm for optimal sensor placement that allows one to find the number, types, and locations of sensors satisfying inhomogeneous coverage requirements and minimizing a specified cost function. The cost function can reflect the actual cost of sensors or other disincentives, e.g., the number of sensors, vulnerability, or emplacement costs of the sensors. The sensors are characterized in terms of a probability of detection, which takes into account signature propagation effects, such as geometrical spreading and inhomogeneous attenuation. The proposed approach incorporates many realistic requirements, e.g., existence of high-value objects, obstacles, forbidden emplacement areas, and perimeter protection. For large spatial grids, the strict optimal solution is, in general, difficult to calculate. A fast algorithm for finding a suboptimal but nonetheless highly satisfactory solution is developed. The developed algorithm is compared against a heuristic algorithm that places sensors one-by-one in the most poorly covered spots. Numerical simulations suggest that the algorithm for a suboptimal solution always outperforms the heuristic algorithm. Software for optimal sensor placement is presented and discussed.

Time-delay estimation (TDE) is a common requirement of the ranging and localization systems often found in unattended ground sensors. In this paper we consider novel approaches to the TDE problem in a time-warping acoustic environment, such as that encountered when the propagation velocity is not constant-due to wind gusts, for example. An increasing propagation velocity induces a compression of the received signal, while a diminishing velocity dilates the signal. These effects warp the shape of the received signal and can significantly reduce the effectiveness of traditional TDE algorithms This paper presents algorithms and performance bounds for TDE in random velocity environments. We model unknown signal warping as a low-pass random process, which serves as a form of non-additive noise in the time-delay estimation problem. For warping environments, we propose computationally efficient non-parametric algorithms for TDE that significantly outperform traditional time-delay estimators, such as the location of the sample cross-correlation peak. The Cram´er-Rao bound for TDE in a time-warping environment is also presented and used to evaluate the proposed estimators. Simulations demonstrate the bounds and estimator performance for acoustic signals.

The performance and utility of battlefield and homeland security sensors depends on many complicated environmental and mission-related factors. This paper describes a general software design for predicting performance of such sensors. It is intended for application to a wide range of sensing modalities and based on an object-oriented framework that can be incorporated into Army command and control (C2) systems, decision support tools (DSTs), and force-on-force simulations. The approach breaks down sensor performance prediction into the following steps: (1) information gathering and construction of the tactical and environmental scenario, (2) translation of the scenario information, (3) target and noise signature prediction models, (4) prediction of sensor performance metrics, and (5) display of and interaction with the information. The main components for Steps 3 and 4 involve operations on signature features, which are described statistically by signal-model objects. These are the units of information needed by the sensor platforms for producing inferences such as the presence or location of a target. The features are generated by emitter platforms, propagated through the environment by feature transmitters (which use scenario translators, Step 2, to convert the atmospheric and terrain descriptions to the necessary model parameters), and then processed by the sensor platforms. To avoid generation and transmission of unneeded data, the architecture is based on a data "pull" (request) from the sensor platforms rather than the more commonly used approach in DSTs of data "push" from the emitter platforms.

Unattended autonomous systems of the future will involve groups of static and mobile sensors functioning in coordination to achieve overall task objectives. Such systems can be viewed as wirelessly networked unmanned heterogeneous sensor networks. We discuss a distributed heterogeneous sensing system with static sensors and mobile robots with novel control optimization algorithms for dynamic adaptation, coordinated control and end to end resource management of all sensors in response to detected events to achieve overall system goals and objectives. Our system is designed for a host of applications, such as unmediated data monitoring and record keeping of the environment, battlefield monitoring using integrated ground, ocean and air sensors, and reactive operation to threats or changing conditions, and homeland security or border/road surveillance systems where unmanned vehicles can be deployed autonomously in response to detected events. Results for large area coastal monitoring are presented. Offline results using actual modeled data from in-situ sensory measurements demonstrate how the sensor parameters can be adapted to maximize observability of a freshwater plume while ensuring that individual system components operate within their physical limitations.^{1 2}

Acoustic signals are a principal detection modality for unattended sensor systems. However, the performance of these systems is frequently suboptimal due to insufficient dynamic range in small systems or excess power consumption in larger systems. This paper discusses an approach to developing an unattended ground sensor (UGS) system that has the best features of both worlds. This system, developed by McQ Inc., has exceptional dynamic range (> 100 dB) while operating at power levels of 1.5-5 watts. The system also has a user definable signal parameter library and automated detection methodology that will be described.

SARA, Inc. has developed microphone arrays that are as effective at reducing flow noise as foam windscreens and sufficiently rugged for tough battlefield environments. These flow noise reducing (FNR) sensors have a metal body and are flat and conformally mounted so they can be attached to the roofs of land vehicles and are resistant to scrapes from branches. Flow noise at low Mach numbers is created by turbulent eddies moving with the fluid flow and inducing pressure variations on microphones. Our FNR sensors average the pressure over the diameter (~20 cm) of their apertures, reducing the noise created by all but the very largest eddies. This is in contrast to the acoustic wave which has negligible variation over the aperture at the frequencies of interest (f less or equal than 400 Hz). We have also post-processed the signals to further reduce the flow noise. Two microphones separated along the flow direction exhibit highly correlated noise. The time shift of the correlation corresponds to the time for the eddies in the flow to travel between the microphones. We have created linear microphone arrays parallel to the flow and have reduced flow noise as much as 10 to 15 dB by subtracting time-shifted signals.

Textron Systems (Textron) has been using geophones for target detection for many years. This sensing capability was utilized for detection and classification purposes only. Recently Textron has been evaluating multiaxis geophones to calculate bearings and track targets more specifically personnel. This capability will not only aid the system in locating personnel in bearing space or cartesian space but also enhance detection and reduce false alarms. Textron has been involved in the testing and evaluation of several sensors at multiple sites. One of the challenges of calculating seismic bearing is an adequate signal to noise ratio. The sensor signal to noise ratio is a function of sensor coupling to the ground, seismic propagation and range to target. The goals of testing at multiple sites are to gain a good understanding of the maximum and minimum ranges for bearing and detection and to exploit that information to tailor sensor system emplacement to achieve desired performance. Test sites include 10A Site Devens, MA, McKenna Airfield Ft. Benning, GA and Yuma Proving Ground Yuma, AZ. Geophone sensors evaluated include a 28 Hz triax spike, a 15 Hz triax spike and a hybrid triax spike consisting of a 10 Hz vertical geophone and two 28 Hz horizontal geophones. The algorithm uses raw seismic data to calculate the bearings. All evaluated sensors have triaxial geophone configuration mounted to a spike housing/fixture. The suite of sensors also compares various types of geophones to evaluate benefits in lower bandwidth. The data products of these tests include raw geophone signals, seismic features, seismic bearings, seismic detection and GPS position truth data. The analyses produce Probability of Detection vs range, bearing accuracy vs range, and seismic feature level vs range. These analysis products are compared across test sites and sensor types.

McQ has developed a miniaturized, programmable, ruggedized data collector intended for use in weapon testing or data collection exercises that impose severe stresses on devices under test. The recorder is designed to survive these stresses which include acceleration and shock levels up to 100,000 G. The collector acquires and stores up to four channels of signal data to nonvolatile memory for later retrieval by a user. It is small (< 7 in³), light weight (< 1 lb), and can operate from various battery chemistries. A built-in menuing system, accessible via a USB interface, allows the user to configure parameters of the recorder operation, such as channel gain, filtering, and signal offsets, and also to retrieve recorded data for analysis. An overview of the collector, its features, performance, and potential uses, is presented.

This paper presents initial object profile classification results using range and elevation independent features from a simulated infrared profiling sensor. The passive infrared profiling sensor was simulated using a LWIR camera. A field data collection effort to yield profiles of humans and animals is reported. Range and elevation independent features based on height and width of the objects were extracted from profiles. The profile features were then used to train and test four classification algorithms to classify objects as humans or animals. The performance of Naïve Bayesian (NB), Naïve Bayesian with Linear Discriminant Analysis (LDA+NB), K-Nearest Neighbors (K-NN), and Support Vector Machines (SVM) are compared based on their classification accuracy. Results indicate that for our data set SVM and (LDA+NB) are capable of providing classification rates as high as 98.5%. For perimeter security applications where misclassification of humans as animals (true negatives) needs to be avoided, SVM and NB provide true negative rates of 0% while maintaining overall classification rates of over 95%.

The Long Wave Infrared (LWIR) Profile Feature Extractor (PFx) sensor has evolved from the initial profiling sensor that was developed by the University of Memphis (Near IR) and the Army Research Laboratory (visible). This paper presents the initial signatures of the LWIR PFx for human with and without backpacks, human with animal (dog), and a number of other animals. The current version of the LWIR PFx sensor is a diverging optical tripwire sensor. The LWIR PFx signatures are compared to the signatures of the Profile Sensor in the visible and Near IR spectral regions. The LWIR PFx signatures were collected with two different un-cooled micro bolometer focal plane array cameras, where the individual pixels were used as stand alone detectors (a non imaging sensor). This approach results in a completely passive, much lower bandwidth, much longer battery life, low weight, small volume sensor that provides sufficient information to classify objects into human Vs non human categories with a 98.5% accuracy.

This paper introduces new developed unattended ground sensor - UGS for urban warfare conditions. It describes the challenges of urban area warfare and the problems rising with standard UGS installed, particularly for data transfer via VHF communication. Then the author discusses the options for UGS new types to be made considering data transfer in LF-band. However, the core deals with the design and construction of an ultra-thin sensor whose part sensing the mechanical vibrations of surface seismic waves excited by human walk is also utilized as a transmitting aerial operating in LF-band. In this part, the paper describes the modified LF-band receiver working at the same time as a UHF-band repeater unit, too. In addition, the receiver may be completed with a GPS-operating receiver unit. The paper concludes with presentation of practical test run results achieved with a sensor sample really made and with a description of future development trends of ultra-thin sensors. The designed sensor may be optionally utilized for non-standard locating unit to find persons' positions in underground spaces, tunnels, caves, etc.

A low cost, lightweight, easily deployable imaging sensor that can dependably discriminate threats from other activities within its field of view and, only then, alert the distant duty officer by transmitting a visual confirmation of the threat would provide a valuable asset to modern defense. At present, current solutions suffer from a multitude of deficiencies - size, cost, power endurance, but most notably, an inability to assess an image and conclude that it contains a threat. The human attention span cannot maintain critical surveillance over banks of displays constantly conveying such images from the field. DigitalTripwire is a small, self-contained, automated human-detection system capable of running for 1-5 days on two AA batteries. To achieve such long endurance, the DigitalTripwire system utilizes an FPGA designed with sleep functionality. The system uses robust vision algorithms, such as a partially unsupervised innovative backgroundmodeling algorithm, which employ several data reduction strategies to operate in real-time, and achieve high detection rates. When it detects human activity, either mounted or dismounted, it sends an alert including images to notify the command center. In this paper, we describe the hardware and software design of the DigitalTripwire system. In addition, we provide detection and false alarm rates across several challenging data sets demonstrating the performance of the vision algorithms in autonomously analyzing the video stream and classifying moving objects into four primary categories - dismounted human, vehicle, non-human, or unknown. Performance results across several challenging data sets are provided.

This paper describes experiments and analysis of seismic signals in addressing the problem of personnel detection for indoor surveillance. Data was collected using geophones to detect footsteps from walking and running in indoor environments such as hallways. Our analysis of the data shows the significant presence of nonlinearity, when tested using the surrogate data method. This necessitates the need for novel detector designs that are not based on linearity assumptions. We present one such method based on empirical mode decomposition (EMD) and functional data analysis (FDA) and evaluate its applicability on our collected dataset.

This paper discusses the development of an Unattended Ground Sensor based on an array of pressure sensors designed to be buried in the ground. This sensor array, along with the required software (still under development), will have the ability to distinguish between humans and animals based on the size and shape of the foot print. The technology may also be applied to determine the weight and type of vehicle traveling on a road. The sensor array consists of pressure sensitive resistors (piezoresistors) on 0.8 inch centers printed on a sheet of polyimide film. Although very large arrays might one day be screen printed, the arrays for this study have been printed using a syringe dispenser and a precision x-y computer controlled table. For the preliminary development, the array has been sized to 8X10 inches. The piezoresistive properties of the sensors are discussed and preliminary test data is presented. It is shown that the piezoresistive gauge factor (ΔR/R/ΔL/L) is roughly 10 times that of conventional metal strain gauges. Because the change in resistance is large compared to metal strain gauges, lower cost electronics can be used. The small net size and low mass enables sensing elements with fast response time. The fact that these piezoresistive elements are directly printed, as opposed to being adhesively attached to a surface, eliminates many of the issues associated with bonded discrete sensors. It is anticipated that the piezoresistive sensor approach presented in this paper will be well suited to extremely rugged environmental conditions compared to the commercially available sensor arrays which rely on surface contact resistance or capacitive sensors which can be easily destroyed by moisture. Environmental testing will be done in a future phase of the project. The final system, which is still under development, will consist of a sensor array, information processing, and RF signal transmission. The system is anticipated to be low cost and environmentally rugged.

The diverse sensor types and networking technologies commonly used in fielded sensro networks provide a unique set of challenges ^[1] in the areas of sensor identification, interoperability, and sensor data consumability. The ITA Senor Fabric is a middleware infrastructure - developed as part of the International Technology Alliance (ITA)^[2] in Network and Information Science - that addresses these challenges by providing unified access to, and management of, sensor networks. The Fabric spans the network from command and control, through forward operating bases, and out to mobile forces and fielded sensors, maximizing the availability and utility of intelligence information to users.

The Current Force Unattended Ground Sensors (UGS) comprise the OmniSense, Scorpion, and Silent Watch systems. As deployed by U.S. Army Central Command in 2006, sensor reports from the three systems were integrated into a common Graphical User Interface (GUI), with three separate vendor-specific applications for Command-and-Control (C2) functions. This paper describes the requirements, system architecture, implementation, and testing of an upgrade to the Processing, Exploitation, and Dissemination back-end server to incorporate common remote Command-and-Control capabilities.