In defense-related robotics research, a mission level integration gap exists between mission tasks (tactical) performed by ground, sea, or air applications and elementary behaviors enacted by processing, communications, sensors, and weaponry resources (platform specific). The gap spans ensemble (heterogeneous team) behaviors, automatic MOE/MOP tracking, and tactical task modeling/simulation for virtual and mixed teams comprised of robotic and human combatants. This study surveys robotic system architectures, compares approaches for navigating problem/state spaces by autonomous systems, describes an architecture for an integrated, repository-based modeling, simulation, and execution environment, and outlines a multi-tiered scheme for robotic behavior components that is agent-based, platform-independent, and extendable via plug-ins. Tools for this integrated environment, along with a distributed agent framework for collaborative task performance are being developed by a U.S. Army funded SBIR project (RDECOM Contract N61339-04-C-0005).

This paper will describe how the Mobility Open Architecture Tools and Simulation (MOAST) framework can facilitate performance evaluations of RCS compliant multi-vehicle autonomous systems. This framework provides an environment that allows for simulated and real architectural components to function seamlessly together. By providing repeatable environmental conditions, this framework allows for the development of individual components as well as component performance metrics. MOAST is composed of high-fidelity and low-fidelity simulation systems, a detailed model of real-world terrain, actual hardware components, a central knowledge repository, and architectural glue to tie all of the components together. This paper will describe the framework’s components in detail and provide an example that illustrates how the framework can be utilized to develop and evaluate a single architectural component through the use of repeatable trials and experimentation that includes both virtual and real components functioning together

A number of researchers have attempted to model human driving or flying skills (e.g., acceleration, steering, vehicle following, etc) in an effort to develop robotic or simulated driver models. In these applications, validation consists of comparing the source data to the model output; thus, it is assumed that model fidelity is correlated with similarity to true human performance. This paper reviews some of the validation metrics found in the literature and discusses the limitations of these metrics. It also presents an alternative metric designed to mitigate these limitations and the test-bed designed to derive this metric. Finally, it illustrates the application of this metric to a number of different trajectory models built through a variety of modeling techniques (e.g., Kalman filters, neural networks, and Newtonian equations).

The performance evaluation of an obstacle detection and segmentation algorithm for Automated Guided Vehicle (AGV) navigation in factory-like environments using a new 3D real-time range camera is the subject of this paper. Our approach expands on the US ASME B56.5 Safety Standard, which now allows for non-contact safety sensors, by performing tests on objects specifically sized in both the US and the British Safety Standards. These successful tests placed the recommended, as well as smaller, material-covered and sized objects on the vehicle path for static measurement. The segmented (mapped) obstacles were then verified in range to the objects and object size using simultaneous, absolute measurements obtained using a relatively accurate 2D scanning laser rangefinder. These 3D range cameras are expected to be relatively inexpensive and used indoors and possibly used outdoors for a vast amount of mobile robot applications building on experimental results explained in this paper.

The level of automation in combat vehicles being developed for the Army's objective force is greatly increased over the Army's legacy force. This automation is taking many forms in emerging vehicles; varying from operator decision aides to fully autonomous unmanned systems. The development of these intelligent vehicles requires a thorough understanding of all of the intelligent behavior that needs to be exhibited by the system so that designers can allocate functionality to humans and/or machines. Traditional system specification techniques focused heavily on the functional description of the major systems and implicitly assumed that a well-trained crew would operate these systems in a manner to accomplish the tactical mission assigned to the vehicle. In order to allocate some or all of these intelligent behaviors to machines in future vehicles it is necessary to be able to identify and describe these intelligent behaviors in detail. In this paper, we describe an effort to develop an intelligent systems (IS) ontology using Protege. The goal of this effort is to develop a common, implementation-independent, extendable knowledge source for researchers and developers in the intelligent vehicle community that will: * Provide a standard set of domain concepts along with their attributes and inter-relations * Allow for knowledge capture and reuse * Facilitate systems specification, design, and integration, and * Accelerate research in the field. This paper describes the methodology we have used to identify knowledge in this domain and an approach to capture and visualize the knowledge in the ontology.

As the Army transforms into the Future Force, particular attention must be paid to operations in Complex and Urban Terrain. Our adversaries increasingly draw us into operations in the urban environment and one can presume that this trend will continue in future battles. In order to ensure that the United States Army maintains battlefield dominance, the Army Research Laboratory (ARL) is developing technology to equip our soldiers for the urban operations of the future. Sophisticated soldier borne systems will extend sensing to the individual soldier, and correspondingly, allow the soldier to establish an accurate picture of their surrounding environment utilizing information from local and remote assets. Robotic platforms will be an integral part of the future combat team. These platforms will augment the team with remote sensing modalities, task execution capabilities, and enhanced communication systems. To effectively utilize the products provided by each of these systems, collected data must be exchanged in real time to all affected entities. Therefore, the Army Research Laboratory is also developing the technology that will be required to support high bandwidth mobile communication in urban environments. This technology incorporates robotic systems that will allow connectivity in areas unreachable by traditional systems. This paper will address some of the issues of providing wireless connectivity in complex and urban terrain. It will further discuss approaches developed by the Army Research Laboratory to integrate communications capabilities into soldier and robotic systems and provide seamless connectivity between the elements of a combat team, and higher echelons.

Teams of robots or mixed teams of warfighters and robots on reconnaissance and other missions can benefit greatly from a local fusion station. A local fusion station is defined here as a small mobile processor with interfaces to enable the ingestion of multiple heterogeneous sensor data and information streams, including blue force tracking data. These data streams are fused and integrated with contextual information (terrain features, weather, maps, dynamic background features, etc.), and displayed or processed to provide real time situational awareness to the robot controller or to the robots themselves. These blue and red force fusion applications remove redundancies, lessen ambiguities, correlate, aggregate, and integrate sensor information with context such as high resolution terrain. Applications such as safety, team behavior, asset control, training, pattern analysis, etc. can be generated or enhanced by these fusion stations. This local fusion station should also enable the interaction between these local units and a global information world.

Spatial data is of crucial importance to situation awareness in the battlefield. As geo-located data from local sensors and global intelligence is gathered and made available, it becomes necessary to route, filter, analyze and present information for decision-making. GIS provides the most intuitive platform for manual/visual correlation of all the gathered information. Too often, presentation applications are tightly coupled with a GIS which limits their use of new GIS products as they emerge. This paper presents a technique for abstracting client applications from the underlying map display to accommodate the many different GIS applications used by military programs.

We are currently developing autonomous urban navigation capabilities for the iRobot PackBot. The TARDEC-funded Wayfarer Project is developing a modular navigation payload that incorporates LIDAR, vision, FLIR, and inertial navigation sensors. This payload can be attached to any PackBot and will provide the robot with the capability to perform fully-autonomous urban reconnaissance missions. These capabilities will enable the PackBot Wayfarer to scout unknown territory and return maps along with video and FLIR image sequences. The Wayfarer navigation payload includes software components for obstacle avoidance, perimeter and street following, and map-building. The obstacle avoidance system enables the PackBot to avoid collisions with a wide range of obstacles in both outdoor and indoor environments. This system combines 360-degree planar LIDAR range data with 3D obstacle detection using stereo vision using a Scaled Vector Field Histogram algorithm. We use a real-time Hough transform to detect linear features in the range data that correspond to building walls and street orientations. We use the LIDAR range data to build an occupancy grid map of the robot's surroundings in real-time. Data from the range sensors, obstacle avoidance, and the Hough transform are transmitted via UDP over wireless Ethernet to an OpenGL-based OCU that displays this information graphically and in real-time.

As the Army transforms to the Future Force, particular attention must be paid to operations in Complex and Urban Terrain. Because our adversaries realize that we don't have battlefield dominance in the urban environment, and because population growth and migration to urban environments is still on the increase, our adversaries will continue to draw us into operations in the urban environment. The Army Research Laboratory (ARL) is developing technology to equip our soldiers for the urban operations of the future. Sophisticated small robotic platforms with diverse sensor suites will be an integral part of the Future Force, and must be able to collaborate not only amongst themselves but also with their manned partners. The use of acoustic sensors on robotic platforms, as shown in this paper, will greatly aid the soldiers of the future force in performing numerous types of missions including Reconnaissance, Surveillance, and Target Acquisition (RSTA) by providing situational awareness, particularly to the dismounted soldier operating in the urban environment. The work conducted by the Army Research Laboratory, discussed in this paper will be transitioned to the FCS-Small Unattended Ground Vehicle (SUGV) program and FFW. The Army Research Laboratory is already working with these programs to ensure a feasible migration path. This paper focuses on four areas relating to acoustic sensing on robots for the urban environment as demonstrated at the DoD Horizontal Fusion Portfolio’s Warriors Edge (WE) Quantum Leap II (QL II) demonstration at Ft Benning, GA in August, 2004: small (man-portable) robot detection, mule-sized robot detection, sensor fusion across multiple platforms, and soldier/robot team interaction.

The Robotic Systems Pool (RSP), sponsored by the Joint Robotics Program (JRP), is an inventory of small robotic systems, payloads, and components intended to expedite the development and integration of technology into effective, supportable, fielded robotic assets. The RSP loans systems to multiple users including the military, first-responders, research organizations, and academia. These users provide feedback in their specific domain, accelerating research and development improvements of robotic systems, which in turn allow the joint warfighter to benefit from such changes more quickly than from traditional acquisition cycles. Over the past year, RSP assets have been used extensively for pre-deployment operator and field training of joint Explosive Ordnance Disposal (EOD) teams, and for the training of Navy Reservist repair technicians. These Reservists are part of the Robotic Systems Combat Support Platoon (RSCSP), attached to Space and Naval Warfare Systems Center, San Diego. The RSCSP maintains and repairs RSP assets and provides deployable technical support for users of robotic systems. Currently, a small team from the RSCSP is deployed at Camp Victory repairing and maintaining man-portable unmanned ground vehicles (UGVs) used by joint EOD teams in Operation Iraqi Freedom. The focus of this paper is to elaborate on the RSP and RSCSP and their role as invaluable resources for spiral development in the robotics community by gaining first-hand technical feedback from the warfighter and other users.

The wars in Iraq and Afghanistan have shown the importance of robotic technology as a force multiplier and a tool for moving soldiers out of harms way. Situations on the ground make soldiers performing checkpoint operations easy targets for snipers and suicide bombers. Robotics technology reduces risk to soldiers and other personnel at checkpoints. Early user involvement in innovative and aggressive development and acquisition strategies are the key to moving robotic and associated technology into the hands of the user. This paper updates activity associated with rapid development of the Omni-Directional Inspection System (ODIS) robot for under vehicle inspection and reports on our field experience with robotics in Iraq and Afghanistan. In February of 2004, two TARDEC Engineers departed for a mission to Iraq and Afghanistan with ten ODIS Robots. Six robots were deployed in the Green Zone in Baghdad. Two Robots were deployed at Kandahar Army Airfield and two were deployed at Bagram Army Airfield in Afghanistan. The TARDEC Engineers who performed this mission trained the soldiers and provided initial on site support. They also trained Exponent employees assigned to the Rapid Equipping Force in ODIS repair. We will discuss our initial deployment, lessons learned and future plans.

The continuing military operations in Iraq and Afghanistan have resulted in a rapidly growing demand for mobile robots to be used during Explosive Ordnance Disposal operations. These robots are predominately used by EOD technicians for surveillance and neutralization of explosive threats from a safe standoff distance. The hazardous nature of the mission these vehicles help perform requires them to be expendable. Current commercially available systems, however, although capable of performing the mission, are costly and are not currently available in the large quantities needed by EOD technicians. The Naval EOD Technology Division (NAVEODTECHDIV) proposes an alternative; a low cost, mobile robot using Commercial Off-The-Shelf (COTS) parts that is specifically tailored to perform hazardous EOD missions. The main functions of this robot are efficient surveillance and explosive threat neutralization. The use of COTS parts allows for streamlined field supportability and repair. A proposed speed of five miles per hour is a drastic improvement over many existing EOD robots and will allow EOD teams to quickly survey and assess potentially dangerous situations. The manipulator will be capable of precision placement of neutralization charges. The cost of this proposed robot is $10,000. Current commercial robots capable of performing these EOD tasks range in price from $40,000 to over $150,000. This conference paper will describe the robot design and prototyping process, from gathering requirements to fabrication and testing.

This paper describes how the functionality of the DoD Joint Architecture for Unmanned Systems (JAUS) was expanded by developing experimental components related to off-line and reactive path planning and world modeling. JAUS defines a set of reusable components and their interfaces. In order to ensure that the JAUS architecture will be applicable to the entire domain of unmanned mobile systems, the following four characteristics were considered throughout its development: vehicle platform independence, mission isolation, hardware independence, and technology independence. The new experimental components described in this paper have these same characteristics. Existing JAUS components readily allow for autonomous path tracking. In this work the JAUS Version 3.0 Global Path Segment Driver, Global Pose Sensor, Velocity State Sensor, and Primitive Driver components were utilized for that task. What is described in this paper is the development of the following new experimental components: World Model, Sensor Arbiter, and Path Manager. The functionality and interface messages of each of these components are presented, followed by a discussion of the performance of the overall system.

Unmanned ground and air systems operating in collaboration have the potential to provide future Joint Forces a significant capability for operations in complex terrain. Ground and air collaborative engagements potentially offer force conservation, perform timely acquisition and dissemination of essential combat information, and can eliminate high value and time critical targets. These engagements can also add considerably to force survivability by reducing soldier and equipment exposure during critical operations. The Office of the Secretary of Defense, Joint Robotics Program (JRP) sponsored Collaborative Engagement Experiment (CEE) is a consolidation of separate Air Force, Army and Navy collaborative efforts to provide a Joint capability. The Air Force Research Laboratory (AFRL), Material and Manufacturing Directorate, Aerospace Expeditionary Force Division, Force Protection Branch (AFRL\MLQF), The Army Aviation and Missile Research, Development and Engineering Center (AMRDEC) Joint Technology Center (JTC)/Systems Integration Laboratory (SIL), and the Space and Naval Warfare Systems Center-San Diego (SSC San Diego) are conducting technical research and proof of principle for an envisioned operational concept for extended range, three dimensional, collaborative operations between unmanned systems, with enhanced situational awareness for lethal operations in complex terrain. This program will assess information requirements and conduct experiments to identify and resolve technical risks for collaborative engagements using Unmanned Ground Vehicles (UGVs) and Unmanned Aerial Vehicles (UAVs). It will research, develop and physically integrate multiple unmanned systems and conduct live collaborative experiments. Modeling and Simulation systems will be upgraded to reflect engineering fidelity levels to greater understand technical challenges to operate as a team. This paper will provide an update of a multi-year program and will concentrate primarily on the JTC/SIL efforts. Other papers will outline in detail the Air Force and Navy portions of this effort.

This paper describes the development and performance of a sensor system that was utilized for autonomous navigation of an unmanned ground vehicle. Four different sensor types were integrated to identify obstacles in the vicinity of the vehicle and to identify smooth terrain that could be traversed at speeds up to thirty miles per hour. The paper also describes a sensor fusion approach that was developed whereby the output of all sensors was in a common grid based format. The environment around the vehicle was modeled by a 120×120 grid where each grid cell was 0.5m× 0.5m in size and where the orientation of the grid lines was always maintained parallel to the north-south and east-west lines. Every sensor output an estimate of the traversability of each grid cell. For the three dimensional obstacle avoidance sensors (rotating ladar and stereo vision) the three dimensional point data was projected onto the grid plane. The terrain traversability sensors, i.e. fixed ladar and monocular vision, estimated traversability based on smoothness of the spatial plane fitted to the range data or the commonality in appearance of pixels in the grid cell to those directly in front of the vehicle, respectively.

Robotic navigation requires that the robotic platform have an idea of its location and orientation within the environment. This localization is known as pose estimation, and has been a much researched topic. There are currently two main categories of pose estimation techniques: pose from hardware, and pose from video (PfV). Hardware pose estimation utilizes specialized hardware such as Global Positioning Systems (GPS) and Inertial Navigation Systems (INS) to estimate the position and orientation of the platform at the specified times. PfV systems use video cameras to estimate the pose of the system by calculating the inter-frame motion of the camera from features present in the images. These pose estimation systems are readily integrated, and can be used to augment and/or supplant each other according to the needs of the application. Both pose from video and hardware pose estimation have their uses, but each also has its degenerate cases in which they fail to provide reliable data. Hardware solutions can provide extremely accurate data, but are usually quite pricey and can be restrictive in their environments of operation. Pose from video solutions can be implemented with low-cost off-the-shelf components, but the accuracy of the PfV results can be degraded by noisy imagery, ambiguity in the feature matching process, and moving objects. This paper attempts to evaluate the cost/benefit comparison between pose from video and hardware pose estimation experimentally, and to provide a guide as to which systems should be used under certain scenarios.

The focus of this paper is on the reconstruction of 3D representations of real world scenes and objects using multiple sensors with, as one of its main applications, the enhancement of the autonomy and mobility of unmanned vehicles. The sensors considered are primarily range acquisition devices (such as laser scanners and stereo systems) that allow the recovery of 3D geometry. One of the most important technical challenges that we are addressing is the registration task in both its multi-modal and single modality aspects. Our work is based on a unified approach that formulates the correspondence problem as an optimization task. In this context we developed a criterion that can be used for 3D free-form shape registration. The new criterion is derived from simple Boolean matching principles by approximation and relaxation. Technically, one of the main advantages of the proposed approach is convexity in the neighborhood of the alignment parameters and continuous differentiability, allowing for the use of standard gradient-based optimization techniques. The proposed algorithm allows also for a significant automation of the scene modeling task by reducing the intervention of human operators in the tedious image registration task. Furthermore, we show that the criterion can be computed in linear time complexity which permits the fast implementation critical in many applications of autonomous mobile platforms.

In this paper, a mobile scanning system for real-time under-vehicle inspection is presented, which is founded on a "Brick" architecture. In this "Brick" architecture, the inspection system is basically decomposed into bricks of three kinds: sensing, mobility, and computing. These bricks are physically and logically independent and communicate with each other by wireless communication. Each brick is mainly composed by five modules: data acquisition, data processing, data transmission, power, and self-management. These five modules can be further decomposed into submodules where the function and the interface are well-defined. Based on this architecture, the system is built by four bricks: two sensing bricks consisting of a range scanner and a line CCD, one mobility brick, and one computing brick. The sensing bricks capture geometric data and texture data of the under-vehicle scene, while the mobility brick provides positioning data along the motion path. Data of these three modalities are transmitted to the computing brick where they are fused and reconstruct a 3D under-vehicle model for visualization and danger inspection. This system has been successfully used in several military applications and proved to be an effective safer method for national security.

The typical algorithm for robot autonomous navigation in off-road complex environments involves building a 3D map of the robot's surrounding environment using a 3D sensing modality such as stereo vision or active laser scanning, and generating an instantaneous plan to navigate around hazards. Although there has been steady progress using these methods, these systems suffer from several limitations that cannot be overcome with 3D sensing and planning alone. Geometric sensing alone has no ability to distinguish between compressible and non-compressible materials. As a result, these systems have difficulty in heavily vegetated environments and require sensitivity adjustments across different terrain types. On the planning side, these systems have no ability to learn from their mistakes and avoid problematic environmental situations on subsequent encounters. We have implemented an adaptive terrain classification system based on the Artificial Immune System (AIS) computational model, which is loosely based on the biological immune system, that combines various forms of imaging sensor inputs to produce a "feature labeled" image of the scene categorizing areas as benign or detrimental for autonomous robot navigation. Because of the qualities of the AIS computation model, the resulting system will be able to learn and adapt on its own through interaction with the environment by modifying its interpretation of the sensor data. The feature labeled results from the AIS analysis are inserted into a map and can then be used by a planner to generate a safe route to a goal point. The coupling of diverse visual cues with the malleable AIS computational model will lead to autonomous robotic ground vehicles that require less human intervention for deployment in novel environments and more robust operation as a result of the system's ability to improve its performance through interaction with the environment.

The Autonomous Intelligent Systems program at Defence R&D Canada-Suffield envisions autonomous systems contributing to decisive operations in the urban battle space. Creating effective intelligence for these systems demands advances in perception, world representation, navigation, and learning. In the land environment, these scientific areas have garnered much attention, while largely ignoring the problem of locomotion in complex terrain. This is a gap in robotics research, where sophisticated algorithms are needed to coordinate and control robotic locomotion in unknown, highly complex environments. Unlike traditional control problems, intuitive and systematic control tools for robotic locomotion do not readily exist thus limiting their practical application. This paper addresses the mobility problem for unmanned ground vehicles, defined here as the autonomous maneuverability of unmanned ground vehicles in unknown, highly complex environments. It discusses the progress and future direction of intelligent mobility research at Defence R&D Canada-Suffield and presents the research tools, topics and plans to address this critical research gap.

When people drive off-road, we look at the upcoming terrain and make a variety of judgments-whether to avoid or attempt to cross a patch of terrain, whether to slow down or speed up, etc. We consider a variety of factors including perceived slope, obstacles, resistance, traction, sinkage, roughness, and the limitations of our perception. We judge many of the handling factors based our recollections of driving on other terrain with similar visual appearance. Perception and terrain understanding algorithms with similar capabilities are needed for unmanned ground vehicle (UGV) autonomous mobility. The objective of this research is to begin to develop methods that can be used by a UGV to learn to associate terrain appearance with handling on the terrain. We demonstrate methods to identify models of speed and acceleration as function of throttle command, and power consumption as a function of speed and acceleration using data collected by on-board sensors as the UGV executes test maneuvers. We demonstrate methods to characterize terrain type from visual appearance. We investigate the hypothesis that terrain with different handling characteristics can be discriminated based on visual appearance characterization.

This work addresses the issue of Terrain Classification that can be applied for path planning for an Unmanned Ground Vehicle (UGV) platform. We are interested in classification of features such as rocks, bushes, trees and dirt roads. Currently, the data is acquired from a color camera mounted on the UGV as we can add range data from a second sensor in the future. The classification is accomplished by first, coarse segmenting a frame and then refining the initial segmentations through a convenient user interface. After the first frame, temporal information is exploited to improve the quality of the image segmentation and help classification adapt to changes due to ambient lighting, shadows, and scene changes as the platform moves. The Mean Shift Classifier algorithm provides segmentation of the current frame data. We have tested the above algorithms with four sequence of frames acquired in an environment with terrain representative of the type we expect to see in the field. A comparison of the results from this algorithm was done with accurate manually-segmented (ground-truth) data, for each frame in the sequence.

Most research on off-road mobile robot sensing focuses on obstacle negotiation, path planning, and position estimation. These issues have conventionally been the foremost factors limiting the performance and speeds of mobile robots. Very little attention has been paid to date to the issue of terrain trafficability, that is, the terrain's ability to support vehicular traffic. Yet, trafficability is of great importance if mobile robots are to reach speeds that human-driven vehicles can reach on rugged terrain. For example, it is obvious that the maximal allowable speed for a turn is lower when driving over sand or wet grass than when driving on packed dirt or asphalt. This paper presents our work on automated real-time characterization of terrain with regard to trafficability for small mobile robots. The two proposed methods can be implemented on skid-steer mobile robots and possibly also on tracked mobile robots. The paper also presents experimental results for each of the two implemented methods.

Tactical behaviors for autonomous ground and air vehicles are an area of high interest to the Army. They are critical for the inclusion of robots in the Future Combat System (FCS). Tactical behaviors can be defined at multiple levels: at the Company, Platoon, Section, and Vehicle echelons. They are currently being defined by the Army for the FCS Unit of Action. At all of these echelons, unmanned ground vehicles, unmanned air vehicles, and unattended ground sensors must collaborate with each other and with manned systems. Research being conducted at the National Institute of Standards and Technology (NIST) and sponsored by the Army Research Lab is focused on defining the Four Dimensional Real-time Controls System (4D/RCS) reference model architecture for intelligent systems and developing a software engineering methodology for system design, integration, test and evaluation. This methodology generates detailed design requirements for perception, knowledge representation, decision making, and behavior generation processes that enable complex military tactics to be planned and executed by unmanned ground and air vehicles working in collaboration with manned systems.

Critical to the development of unmanned ground vehicle platforms is the incorporation of adaptive tactical behaviors for the planning of high-level navigation and tactical actions. BBN Technologies recently completed a simulation-based project for the Army Research Lab (ARL) in which we applied an evolutionary computation approach to navigating through a terrain to capture flag objectives while faced with one or more mobile enemies. Our Advocates and Critics for Tactical Behaviors (ACTB) system evolves plans for the vehicle that control its movement goals (in the form of waypoints), and its future actions (e.g., pointing cameras). We apply domain-specific, state-dependent genetic operators called advocates that promote specific tactical behaviors (e.g., adapt a plan to stay closer to walls). We define the fitness function as a weighted sum of a number of independent, domain-specific, state-dependent evaluation components called critics. Critics reward plans based upon specific tactical criteria, such as minimizing risk of exposure or time to the flags. Additionally, the ACTB system provides the capability for a human commander to specify the "rules of engagement" under which the vehicle will operate. The rules of engagement determine the planning emphasis required under different tactical situations (e.g., discovery of an enemy), and provide a mechanism for automatically adapting the relative selection probabilities of the advocates, the weights of the critics, and the depth of planning in response to tactical events. The ACTB system demonstrated highly effective performance in a head-to-head testing event, held by ARL, against two competing tactical behavior systems.

This paper details the development of a minimal set of locally distributed navigation beacons that can provide new waypoints in dense obstacle fields. The 'beacons' provide direction and magnitude inputs for the robot to use for its next waypoint. The beacons are placed in such a manner that all locations within a bounded playing field can reach a goal area in a desired number of steps. This guarantee of total coverage comes only with tuning the magnitudes and directions of each beacon (as well as their position in the field). Key to this approach is the underlying 'color map'. The color map assigns a color to regions of the playing field based on whether the region terminates at the goal ('green'); leaves the playing field and doesn't return ('red'); or doesn’t leave the playing field but does not terminate at the goal (within a fixed number of steps)-also know as 'stagnation' ('yellow'). Changes in the placement of the beacons and their associated parameters result in changes to the color map. A software tool has been developed to allow a user to see the instantaneous changes in the color map when changes are made to the beacons. This paper will also describe how the beacons are related to both Voronoi diagrams and nearest neighbor classifiers-thus generating the final name for the navigation beacons; Voronoi Classifiers. Future work is detailed including the development of color maps for other cost metrics (such as distance traveled, power consumed or terrain trafficabilty) and efforts in developing an algorithm to find the infimum solution (minimize the maximum steps, distance, etc.).

Advances in networking, communications, and sensor technology have made it possible to deploy networks of spatially-distributed sensors to provide tactical or security forces with critical operational intelligence. Many application arenas have requirements for low-power, long-life sensors and stand-off for operations personnel, dictating the need for socalled unattended ground sensor networks (UGS). However, often the environment (e.g., an urban environment) limits communication capabilities. To solve this problem, in this paper we propose the use of multiple, cooperating unmanned aerial vehicles (UAVs) to collect, or exfiltrate, the data from the UGS. In our scenario several prototype sensor clusters are deployed in relatively unknown locations, with a limited communication range. Each cluster's dedicated base/bridge node communicates via an IEEE 802.11b wireless link. The UAVs execute a cooperative flight pattern to first find the sensor clusters, to then collect their data, and to relay that data to a ground station. Throughout the process, the UAVs cooperate autonomously to achieve the goal, reconfiguring as needed in response to changes in the mission, in the targets (sensors), or in the available resources (the UAVs themselves). In the paper we describe our approach to cooperative behavior, our aircraft, and some preliminary experimental results.

We review a large multidisciplinary effort to develop a family of autonomous robots capable of rapid, agile maneuvers in and around natural and artificial vertical terrains such as walls, cliffs, caves, trees and rubble. Our robot designs are inspired by (but not direct copies of) biological climbers such as cockroaches, geckos, and squirrels. We are incorporating advanced materials (e.g., synthetic gecko hairs) into these designs and fabricating them using state of the art rapid prototyping techniques (e.g., shape deposition manufacturing) that permit multiple iterations of design and testing with an effective integration path for the novel materials and components. We are developing novel motion control techniques to support dexterous climbing behaviors that are inspired by neuroethological studies of animals and descended from earlier frameworks that have proven analytically tractable and empirically sound. Our near term behavioral targets call for vertical climbing on soft (e.g., bark) or rough surfaces and for ascents on smooth, hard steep inclines (e.g., 60 degree slopes on metal or glass sheets) at one body length per second.

In this paper, we describe our recent results in the development of a new class of soft, continuous backbone ("continuum") robot manipulators. Our work is strongly motivated by the dexterous appendages found in cephalopods, particularly the arms and suckers of octopus, and the arms and tentacles of squid. Our ongoing investigation of these animals reveals interesting and unexpected functional aspects of their structure and behavior. The arrangement and dynamic operation of muscles and connective tissue observed in the arms of a variety of octopus species motivate the underlying design approach for our soft manipulators. These artificial manipulators feature biomimetic actuators, including artificial muscles based on both electro-active polymers (EAP) and pneumatic (McKibben) muscles. They feature a "clean" continuous backbone design, redundant degrees of freedom, and exhibit significant compliance that provides novel operational capacities during environmental interaction and object manipulation. The unusual compliance and redundant degrees of freedom provide strong potential for application to delicate tasks in cluttered and/or unstructured environments. Our aim is to endow these compliant robotic mechanisms with the diverse and dexterous grasping behavior observed in octopuses. To this end, we are conducting fundamental research into the manipulation tactics, sensory biology, and neural control of octopuses. This work in turn leads to novel approaches to motion planning and operator interfaces for the robots. The paper describes the above efforts, along with the results of our development of a series of continuum tentacle-like robots, demonstrating the unique abilities of biologically-inspired design.

Future human and robotic planetary expeditions could benefit greatly from expanded Extra-Vehicular Activity (EVA) capabilities supporting a broad range of multiple, concurrent surface operations. Risky, expensive and complex, conventional EVAs are restricted in both duration and scope by consumables and available manpower, creating a resource management problem. A mobile, highly dexterous Extra-Vehicular Robotic (EVR) system called Centaur is proposed to cost-effectively augment human astronauts on surface excursions. The Centaur design combines a highly capable wheeled mobility platform with an anthropomorphic upper body mounted on a three degree-of-freedom waist. Able to use many ordinary handheld tools, the robot could conserve EVA hours by relieving humans of many routine inspection and maintenance chores and assisting them in more complex tasks, such as repairing other robots. As an astronaut surrogate, Centaur could take risks unacceptable to humans, respond more quickly to EVA emergencies and work much longer shifts. Though originally conceived as a system for planetary surface exploration, the Centaur concept could easily be adapted for terrestrial military applications such as de-mining, surveillance and other hazardous duties.

This paper describes the design and performance of the OmniTread serpentine robot, developed at the University of Michigan. Serpentine robots are mobile robots that comprise of multiple rigid segments, connected by actuated joints. The segments usually have drive elements, such as wheels or tracks. To date, we have developed two versions of the OmniTread. The larger version, called OT-8, has five rigid segments and four 2-Degree-of-Freedom (2-DOF) joints, and it can drive through an 8-inch diameter opening. The OT-8 is fully functional and this paper documents experimental results for the OT-8. The smaller and newer version, called "OT-4," will have seven segments, six 2- DOF joints, and it will fit through a 4 inch diameter hole. The OT-4 is not yet completely built, but its design is mostly completed and key improvements over the OT-8 have been bench tested. The foremost and unique design characteristic of the OmniTread is the use of pneumatic bellows to actuate the joints. The pneumatic bellows allow the simultaneous control of position and stiffness for each joint. Controllable stiffness is of crucial importance in serpentine robots, which require stiff joints to cross gaps and compliant joints to conform to rough terrain for effective propulsion. Another unique feature of the OmniTread design is the maximal coverage of all four sides of each segment with driven tracks. This design makes the robot indifferent to roll-overs, which are bound to happen when the long and slen-der bodies of serpentine robots travel over rugged terrain.

As the number of the degrees of motion freedom increase in a robotic system, so grows the difficulty of control. We describe a model of a novel highly flexible robotic architecture composed of hundreds of motor elements, each associated with a unique degree of motion freedom. This new robotic architecture possesses a variably compliant structure that allows for the controlled distribution of loads and forces, and for the maintenance of different conformations. We then suggest two methods of intelligent control to manage the many motor elements. One method derives from neural networks, the other involves algorithms inspired by the biological immune system. Both methods are based on the system's perception of its own kinematics, and later self-prediction of the forces generated by coordinated subsets of motor elements that accomplish robot mobility and other work upon the environment.

Passive perception of terrain features is a vital requirement for military related unmanned autonomous vehicle operations, especially under electromagnetic signature management conditions. As a member of Team Raptor, the Jet Propulsion Laboratory developed a self-contained passive perception system under the DARPA funded PerceptOR program. An environmentally protected forward-looking sensor head was designed and fabricated in-house to straddle an off-the-shelf pan-tilt unit. The sensor head contained three color cameras for multi-baseline daytime stereo ranging, a pair of cooled mid-wave infrared cameras for nighttime stereo ranging, and supporting electronics to synchronize captured imagery. Narrow-baseline stereo provided improved range data density in cluttered terrain, while wide-baseline stereo provided more accurate ranging for operation at higher speeds in relatively open areas. The passive perception system processed stereo images and outputted over a local area network terrain maps containing elevation, terrain type, and detected hazards. A novel software architecture was designed and implemented to distribute the data processing on a 533MHz quad 7410 PowerPC single board computer under the VxWorks real-time operating system. This architecture, which is general enough to operate on N processors, has been subsequently tested on Pentium-based processors under Windows and Linux, and a Sparc based-processor under Unix. The passive perception system was operated during FY04 PerceptOR program evaluations at Fort A. P. Hill, Virginia, and Yuma Proving Ground, Arizona. This paper discusses the Team Raptor passive perception system hardware and software design, implementation, and performance, and describes a road map to faster and improved passive perception.

As part of DARPA's MARS2020 program, the Jet Propulsion Laboratory has developed a vision-based system for localization in urban environments that requires neither GPS nor active sensors. System hardware consists of a pair of small FireWire cameras and a standard Pentium-based computer. The inputs to the software system consist of: 1) a crude grid-based map describing the positions of buildings, 2) an initial estimate of robot location and 3) the video streams produced by the stereo pair. At each step during the traverse the system: captures new image data, finds image features hypothesized to lie on the outside of a building, computes the range to those features, determines an estimate of the robot's motion since the previous step and combines that data with the map to update a probabilistic representation of the robot's location. This probabilistic representation allows the system to simultaneously represent multiple possible locations. For our testing, we have derived the a priori map manually using non-orthorectified overhead imagery, although this process could be automated. The software system consists of three primary components. The first is a stereo-based visual odometry system that calculates the 6-degree of freedom camera motion between sequential frames. The second component uses a set of heuristics to identify straight-line segments that are likely to be part of a building exterior. Ranging to these straight-line features is computed using binocular or wide-baseline stereo. The resulting features and the associated range measurements are fed to the third software component, a particle-filter based localization system. This system uses the map and the most recent results from the first two to update the estimate of the robot's location. This report summarizes the design of both the hardware and software and describes the results of applying the system to the global localization of a camera system over an approximately half-kilometer traverse across JPL's Pasadena campus.

The French Military Robotic Study Program (introduced in Aerosense 2003), sponsored by the French Defense Procurement Agency and managed by Thales as the prime contractor, focuses on about 15 robotic themes which can provide an immediate "operational add-on value". The paper details the "constrained navigation" study (named TEL2), which main goal is to identify and test a well-balanced task sharing between man and machine to accomplish a robotic task that cannot be performed autonomously at the moment because of technological limitations. The chosen function is "obstacle avoidance" on rough ground and quite high speed (40 km/h). State of the art algorithms have been implemented to perform autonomous obstacle avoidance and following of forest borders, using scanner laser sensor and standard localization functions. Such an "obstacle avoidance" function works well most of the time, BUT fails sometimes. The study analyzed how the remote operator can manage such failures so that the system remains fully operationally reliable; he can act according to two ways: a) finely adjust the vehicle current heading; b) take the control of the vehicle "on the fly" (without stopping) and bring it back to autonomous behavior when motion is secured again. The paper also presents the results got from the military acceptance tests performed on French 4x4 DARDS ATD.

Most goal-oriented mobile robot tasks involve navigation to one or more known locations. This is generally done using GPS coordinates and landmarks outdoors, or wall-following and fiducial marks indoors. Such approaches ignore the rich source of navigation information that is already in place for human navigation in all man-made environments: signs. A mobile robot capable of detecting and reading arbitrary signs could be tasked using directions that are intuitive to hu-mans, and it could report its location relative to intuitive landmarks (a street corner, a person's office, etc.). Such ability would not require active marking of the environment and would be functional in the absence of GPS. In this paper we present an updated version of a system we call Sign Understanding in Support of Autonomous Navigation (SUSAN). This system relies on cues common to most signs, the presence of text, vivid color, and compact shape. By not relying on templates, SUSAN can detect a wide variety of signs: traffic signs, street signs, store-name signs, building directories, room signs, etc. In this paper we focus on the text detection capability. We present results summarizing probability of detection and false alarm rate across many scenes containing signs of very different designs and in a variety of lighting conditions.

This paper presents a novel approach to tracking groups of objects within clusters that exhibit swarm type behavior. The algorithm for achieving swarm track enables a real-time local reference frame to be established for members of the group in a fashion that makes hand-over and high-resolution tracking possible. Individual members once tracked in a local reference system are easily described and tagged for exchange with other sensors from disparate vantage points making sensor registration a more tractable problem. We demonstrate this algorithm as a GOTO capability for high-resolution lunar feature tracking to illustrate the fine points and details.

The French Military Robotic Study Program (introduced in Aerosense 2003), sponsored by the French Defense Procurement Agency and managed by Thales as the prime contractor, focuses on about 15 robotic themes which can provide an immediate "operational add-on value". The paper details the "automatic speed adjustment" behavior (named SYR4), developed by Giat Industries Company, which main goal is to secure the teleoperated mobility of high speed tracked vehicles on rough grounds; more precisely, the validated low level behavior continuously adjusts the vehicle speed taking into account the teleperator wish AND the maximum speed that the vehicle can manage safely according to the commanded radius of curvature. The algorithm is based on a realistic physical model of the ground-tracks relation, taking into account many vehicle and ground parameters (such as ground adherence and dynamic specificities of tracked vehicles). It also deals with the teleoperator-machine interface, providing a balanced strategy between both extreme behaviors: a) maximum speed reduction before initiating the commanded curve; b) executing the minimum possible radius without decreasing the commanded speed. The paper presents the results got from the military acceptance tests performed on tracked SYRANO vehicle (French Operational Demonstrator).

Autonomous Solutions has developed Chaos, a small unmanned ground vehicle with four modular running gear receptacles. Running gear attached to the vehicle can include any combination of wheels, tracks, articulated and shape-shifting tracks, and legs. Each unit is independently controllable and field changeable. This modular design allows Chaos to combine the strengths of traditional and bio-inspired locomotion. The vehicle offers unprecedented mobility potential including walking, stair climbing, clambering over obstacles, steep slope traversal and extrication. To fully exploit the vehicle's mobility potential, intelligent behaviors must be integrated to reduce the complexity of vehicle operation. Mobility behaviors include operator assisted tele-operation, adaptive gaits, obstacle characterization, traversal, and extrication. This paper will describe the design and development of Chaos including running gear and intelligent mobility behaviors.

The initial construct of the framework for the Autonomy Levels of Unmanned Systems (ALFUS) was presented in the 2004 SPIE Defense and Security Symposium. This paper describes the continuing development effort and further accomplishments made by the Ad Hoc working group. We focus on two elements of the ALFUS product set, namely, the detailed model that is being implemented as a spreadsheet-based tool and the summary model. We also discuss identified challenges.

This paper will briefly outline major activities in Department of Energy (DOE) Laboratories focused on mobile platforms, both Unmanned Ground Vehicles (UGV’s) as well as Unmanned Air Vehicles (UAV's). The activities will be discussed in the context of the science and technology construct used by the DOE Technology Roadmap for Robotics and Intelligent Machines (RIM)1 published in 1998; namely, Perception, Reasoning, Action, and Integration. The activities to be discussed span from research and development to deployment in field operations. The activities support customers in other agencies. The discussion of "perception" will include hyperspectral sensors, complex patterns discrimination, multisensor fusion and advances in LADAR technologies, including real-world perception. "Reasoning" activities to be covered include cooperative controls, distributed systems, ad-hoc networks, platform-centric intelligence, and adaptable communications. The paper will discuss "action" activities such as advanced mobility and various air and ground platforms. In the RIM construct, "integration" includes the Human-Machine Integration. Accordingly the paper will discuss adjustable autonomy and the collaboration of operator(s) with distributed UGV's and UAV's. Integration also refers to the applications of these technologies into systems to perform operations such as perimeter surveillance, large-area monitoring and reconnaissance. Unique facilities and test beds for advanced mobile systems will be described. Given that this paper is an overview, rather than delve into specific detail in these activities, other more exhaustive references and sources will be cited extensively.

The Defence R&D Canada (DRDC) has been given strategic direction to pursue research to increase the independence and effectiveness of military vehicles and systems. This has led to the creation of the Autonomous Intelligent Systems (AIS) program which is notionally divided into air, land and marine vehicles. This paper presents an overarching description of DRDC's concluding, starting and pending future programs in vehicle intelligence and autonomy.

This presentation will provide program information, goals and objectives of the Technology for Human-Robot Interactions in Soldier-Robot Teaming (HRI) Army Technology Objective (ATO). The intent of this program is to develop and demonstrate an intelligent scaleable interface for mounted and dismounted control of ground and air unmanned systems. Currently in the Army there are unique interfaces developed by engineers for each unmanned system fielded. This saddles the soldier with a training burden to learn specific interface operations prior to controlling the robot. By providing a consistent look and feel across various sized controlling devices, the training burden is reduced as well as the soldier's cognitive workload. Additionally, task analysis will be performed to identify workload barriers and bottlenecks, and intelligent agents will be developed and applied to reduce and/or automate the higher workload tasks. Lastly, this program will develop adaptive automation techniques to intelligently shed or introduce tasks at the appropriate time to the soldier to maintain optimal situational awareness and maximize the performance of the soldier-robot team.

In May 2003, the Federal Republic of Germany and the Republic of France awarded a contract to RHEINMETALL LANDSYSTEME GmbH (Germany), MBDA (France) and THALES (France) for the joint development and manufacturing of a technology demonstrator for a vehicle-based close-in countermine system. The objective of this cooperation project, known as MMSR-SYDERA, is to show that, in a full-scale development program, it will be possible to fulfill the joint operational requirements issued by the German and French armies, which are based on the following missions: Fast route opening, Sensitive route opening and Area Clearing . In order to fulfill the three different missions and deal with an extensive array of mine threats, the MMSR-SYDERA countermine system combines two modes of countermine operation, i.e. triggering mines by decoying or detecting mines with sensors for low-order clearing. Thus, the plan calls for the MMSR-SYDERA system to be composed of five vehicles deployed in different configurations in a convoy on the roads to be cleared. This program takes advantage of the latest technologies to reach the objective performances: some innovative decoy tools already validated during real trials and optimized for MMSRSYDERA program, the latest ambitious technologies for detection and confirmation, and state-of-the-art remote-control capabilities, including tele-operation and semi-autonomous autofollow mode, with the highest established performances. An evaluation of the system demonstrator will be carried out at the beginning of 2007.

In future operations unmanned systems must act in concert with manned vehicles and work as a team to accomplish mission-level objectives with minimal human intervention or control. Current technology requires one or more operators/controllers for each unmanned vehicle, is void of inter-vehicle collaboration, and provides limited true autonomy. The Unmanned Autonomous Collaborative Operation program (UACO) is an Army research and development effort designed to advance the state of the art in these areas and demonstrate militarily relevant mission behaviors involving multiple unmanned air vehicles. Multi-UAV autonomy command and control programs are underway in other elements of the Army, Defense Advanced Research Projects Agency (DARPA), Air Force Research Lab (AFRL), and the Navy; however, most conclude with simulation demonstration. This project builds on such efforts but takes the next step toward maturity by including flight demonstrations of UACO mission behaviors in Army aviation and infantry support contexts.

The mission of the Unmanned Systems Branch of SPAWAR Systems Center, San Diego (SSC San Diego) is to provide network-integrated robotic solutions for Command, Control, Communications, Computers, Intelligence, Surveillance, and Reconnaissance (C4ISR) applications, serving and partnering with industry, academia, and other government agencies. We believe the most important criterion for a successful acquisition program is producing a value-added end product that the warfighter needs, uses and appreciates. Through our accomplishments in the laboratory and field, SSC San Diego has been designated the Center of Excellence for Small Robots by the Office of the Secretary of Defense Joint Robotics Program. This paper covers the background, experience, and collaboration efforts by SSC San Diego to serve as the "Impedance-Matching Transformer" between the robotic user and technical communities. Special attention is given to our Unmanned Systems Technology Imperatives for Research, Development, Testing and Evaluation (RDT&E) of Small Robots. Active projects, past efforts, and architectures are provided as success stories for the Unmanned Systems Development Approach.

The pocket-sized ThrowBot is a sub-kilogram-class robot that provides short-range remote eyes and ears for urban combat. This paper provides an overview of lessons learned from experience, testing, and evaluation of the iRobot ThrowBot developed under the Defense Advanced Research Projects Agency (DARPA) Tactical Mobile Robots (TMR) program. Emphasis has been placed on investigating requirements for the next generation of ThrowBots to be developed by iRobot Corporation and SPAWAR Systems Center, San Diego (SSC San Diego) Unmanned Systems Branch. Details on recent evaluation activities performed at the Military Operations in Urban Terrain (MOUT) test site at Fort Benning, GA, are included, along with insights obtained throughout the development of the ThrowBot since its inception in 1999 as part of the TMR program.

The Technology Transfer project employs a spiral development process to enhance the functionality and autonomy of mobile systems in the Joint Robotics Program (JRP) Robotic Systems Pool (RSP). The approach is to harvest prior and on-going developments that address the technology needs identified by emergent in-theatre requirements and users of the RSP. The component technologies are evaluated on a transition platform to identify the best features of the different approaches, which are then integrated and optimized to work in harmony in a complete solution. The result is an enabling mechanism that continuously capitalizes on state-of-the-art results from the research environment to create a standardized solution that can be easily transitioned to ongoing development programs. This paper focuses on particular research areas, specifically collision avoidance, simultaneous localization and mapping (SLAM), and target-following, and describes the results of their combined integration and optimization over the past 12 months.

Many advances have been made in autonomy for unmanned ground vehicles (UGVs) but most of these advances have been for large UGVs only, in that the sensors required for autonomy are typically large, heavy and require a significant amount of power. Because of the size, weight and power restrictions imposed by a man-portable UGV advances in autonomy have been very limited. The SPAWAR Systems Center San Diego (SSC San Diego) has previously developed a waypoint navigation capability for small robots. That system required an operator to monitor a live video feed from the vehicle to ensure it did not strike any obstacles in its path. Now SSC San Diego in cooperation with the NASA Jet Propulsion Laboratory (JPL) has developed a miniature obstacle detection sensor suitable for small robots. SSC San Diego has also developed the obstacle-avoidance algorithms to navigate autonomously around obstacles.

Robot technology is developing at a rapid rate for both commercial and Department of Defense (DOD) applications. As a result, the task of managing both technology and experience information is growing. In the not-to-distant past, tracking development efforts of robot platforms, subsystems and components was not too difficult, expensive, or time consuming. To do the same today is a significant undertaking. The Mobile Robot Knowledge Base (MRKB) provides the robotics community with a web-accessible, centralized resource for sharing information, experience, and technology to more efficiently and effectively meet the needs of the robot system user. The resource includes searchable information on robot components, subsystems, mission payloads, platforms, and DOD robotics programs. In addition, the MRKB website provides a forum for technology and information transfer within the DOD robotics community and an interface for the Robotic Systems Pool (RSP). The RSP manages a collection of small teleoperated and semi-autonomous robotic platforms, available for loan to DOD and other qualified entities. The objective is to put robots in the hands of users and use the test data and fielding experience to improve robot systems.

The Mobile Detection Assessment Response System (MDARS) provides physical security for Department of Defense bases and depots using autonomous unmanned ground vehicles (UGVs) to patrol the site while operating payloads for intruder detection and assessment, barrier assessment, and product assessment. MDARS is in the System Development and Demonstration acquisition phase and is currently undergoing developmental testing including an Early User Appraisal (EUA) at the Hawthorne Army Depot, Nevada-the world's largest army depot. The Multiple Resource Host Architecture (MRHA) allows the human guard force to command and control several MDARS platforms simultaneously. The MRHA graphically displays video, map, and status for each resource using wireless digital communications for integrated data, video, and audio. Events are prioritized and the user is prompted with audio alerts and text instructions for alarms and warnings. The MRHA also interfaces to remote resources to automate legacy physical devices such as fence gate controls, garage doors, and remote power on/off capability for the MDARS patrol units. This paper provides an overview and history of the MDARS program and control station software with details on the installation and operation at Hawthorne Army Depot, including discussions on scenarios for EUA excursions. Special attention is given to the MDARS technical development strategy for spiral evolutions.

Weapon payloads are becoming increasingly important components of unmanned ground vehicles (UGVs). However weapon payloads are extremely difficult to teleoperate. This paper explores the issues involved with automating several aspects of the operations of a weapon payload. These operations include target detection, acquisition, and tracking. Various approaches to these issues are discussed, and the development and results from two different working prototype systems developed at Space and Naval Warfare Systems Center, San Diego (SSC San Diego) are presented. One approach employs a motion-based scheme for target identification, while the second employs an appearance based scheme. Target selection, arming and firing remain teleoperated in both systems.

Under threat of attack, survivability depends on evasive maneuvers exhibited by a squad of robotic vehicles. Reduced vernerability of the squad derives from interruption and redirection of repeated reactions. Exclusion of predictability in the selection of operational routes and lanes, whether preplanned or determined in-stide, renders swarm counter threats useless.

Unmanned vehicles perform critical mission functions. Today, fielded unmanned vehicles have restricted operations as a single asset controlled by a single operator. In the future, however, it is envisioned that multiple unmanned air, ground, surface and underwater vehicles will be deployed in an integrated unmanned (and "manned") team fashion in order to more effectively execute complex mission scenarios. To successfully facilitate this transition from single platforms to an integrated unmanned system concept, it is essential to first develop the required base technologies for multi-vehicle mission requirements, as well as test and evaluate such technologies in tightly controlled field experiments. Under such conditions, advances in unmanned technologies and associated system configurations can be empirically evaluated and quantitatively measured against relevant performance metrics. A series of field experiments will be conducted for unmanned force protection system applications. A basic teaming scenario is: Unmanned aerial vehicles (UAVs) detect a target of interest on the ground; the UAVs cue unmanned ground vehicles (UGVs) to the area; the UGVs provide on-ground evaluation and assessment; and the team of UAVs and UGVs execute the appropriate level of response. This paper details the scenarios and the technology enablers for experimentation using unmanned protection systems.

Development of Unmanned Ground Vehicles (UGVs) has been ongoing for decades. Much of the technology developed for UGVs can be applied directly to Unmanned Surface Vehicles (USVs) with little or no modification. SPAWAR Systems Center San Diego (SSC San Diego) has successfully demonstrated this by transitioning technology (both hardware and software) from a man-portable UGV to a USV demonstrator platform. By transitioning technology already proven in a UGV, SSC San Diego was able to develop a working USV much more quickly than would have been otherwise possible. The technologies ported from the UGV to the USV include: the software architecture and protocol, teleoperation, a Kalman filter for state estimates, waypoint navigation, the Operator Control Unit (OCU), miniature processors, Ethernet switches and a video CODEC board.

Head-aimed vision systems provide significantly improved situational awareness, accuracy, and decision speed for the tele-operation of agile robots. With head-aimed vision, wherever the operator looks, a sensor system onboard the remote vehicle "looks". When done correctly, head-aimed vision crates a powerful sense of telepresence. An overall performance increase of 250% was documented in tests we ran of reconnaissance tasks for an unmanned ground vehicle. Operator workload was also reduced.

Giat Industries, expert in combat system, designs and integrates the technologies (electrical network, mechanical design, ergonomic and mobility command and control) for the French military and civil robotic. One of the main problems of current robotics is to treat the lack of operator's perception due to teleoperation. The article details the assistance for the steer control part of mobility command and control. The purpose of the assistance is to secure the vehicle, to insure the continuity of the command (in particular for tracked vehicle equipped of hydrostatic steering group), and to adapt the command (to use all the range of the command even at high speed and to adapt the speed to the desired curve). The advantage of the system presented in the article is to filter the order first and to transmit to the actuators only safe orders preserving the vehicle. This approach transforms the telepilot's order watching and correction problem, to a telepilot's order filtering problem. We pass here, from a logical "a posteriori" to a logical of "a priori" processing. The assistance includes: a model of the dynamic behavior of the vehicle and an open loop, some feed forward with appropriate filters and a closed loop. The article presents the result of the command studies from time simulation to "man on the loop" simulation. This system, at present time applied to the French Observation Robot demonstrator SYRANO (developped for the DGA-French Defence Procurement Agency), is the subject of a patent.

In this paper we present a map-based approach to localization. We consider indoor navigation in known environments based on the idea of a "vector cloud" by observing that any point in a building has an associated vector defining its distance to the key structural components (e.g., walls, ceilings, etc.) of the building in any direction. Given a building blueprint we can derive the "ideal" vector cloud at any point in space. Then, given measurements from sensors on the robot we can compare the measured vector cloud to the possible vector clouds cataloged from the blueprint, thus determining location. We present algorithms for implementing this approach to localization, using the Hamming norm, the 1-norm, and the 2-norm. The effectiveness of the approach is verified by experiments on a 2-D testbed using a mobile robot with a 360° laser range-finder and through simulation analysis of robustness.

Our research efforts focus on the deployment of 3D sensing capabilities to a multi-modal under vehicle inspection robot. In this paper, we outline the various design challenges towards the automation of the 3D scene modeling task. We employ laser-based range imaging techniques to extract the geometry of a vehicle's undercarriage and present our results after range integration. We perform shape analysis on the digitized triangle mesh models by segmenting them into smooth surface patches based on the curvedness of the surface. Using a region-growing procedure, we then obtain the patch adjacency. On each of these patches, we apply our definition of the curvature variation measure (CVM) as a descriptor of surface shape complexity. We base the information-theoretic CVM on shape curvature, and extract shape information as the entropic measure of curvature to represent a component as a graph network of patches. The CVM at the nodes of the graph describe the surface patch. We then demonstrate our algorithm with results on automotive components. With apriori manufacturer information about the CAD models in the undercarriage we approach the technical challenge of threat detection with our surface shape description algorithm on the laser scanned geometry.

We have designed a minimally intrusive Operator Control Unit (OCU) intended to be used by a dismounted soldier. The OCU is operated using a combination of head aiming, plus a small wireless controller that is integrated in the grip of the soldier's rifle. This minimally intrusive OCU allows soldiers to navigate a software interface (for example, to call up a map), operate a remote camera system or other sensors on an unmanned vehicle, and/or tele-operate the vehicle itself, all while the soldiers have their heads up and their hands on their weapons. Central to the concept is the idea of a head-aimed software interface, where natural and intuitive head motion is used instead of traditional mouse movements to efficiently navigate, point or even select items in the display-operators simply move their heads in the direction that they want to "look" and the display is seamlessly updated with new information. When combined with the controller that is integrated in the weapon grip, this allows almost hands free operation, as opposed to operating a PDA or other standard controller system which generally occupies both hands and requires operators to look down at a screen.

Over the years, scientists and artists alike have imagined walking mechanisms that mimic the natural gait of humans and animals. Only recently have engineers begun to unravel the mystery of animal locomotion. Several walking robots have been built in the past few years [1]. An ongoing research problem with these robots is their inefficiency. Whereas animal locomotion is quite efficient, our efforts to mimic it have not been, with a few notable exceptions [2]. In this paper, we present a design for efficient legged locomotion, and we show the initial concept demonstration.

Increased threats to gate security have caused recent need for improved vehicle inspection methods at security checkpoints in various fields of defense and security. A fast, reliable system of under-vehicle inspection that detects possibly harmful or unwanted materials hidden on vehicle undercarriages and notifies the user of the presence of these materials while allowing the user a safe standoff distance from the inspection site is desirable. An autonomous under-vehicle inspection system would provide for this. The proposed system would function as follows: A low-clearance tele-operated robotic platform would be equipped with sonar/laser range finding sensors as well as a video camera. As a vehicle to be inspected enters a checkpoint, the robot would autonomously navigate under the vehicle, using algorithms to detect tire locations for weigh points. During this navigation, data would be collected from the sonar/laser range finding hardware. This range data would be used to compile an impression of the vehicle undercarriage. Once this impression is complete, the system would compare it to a database of pre-scanned undercarriage impressions. Based on vehicle makes and models, any variance between the undercarriage being inspected and the impression compared against in the database would be marked as potentially threatening. If such variances exist, the robot would navigate to these locations and place the video camera in such a manner that the location in question can be viewed from a standoff position through a TV monitor. At this time, manual control of the robot navigation and camera control can be taken to imply further, more detailed inspection of the area/materials in question. After-market vehicle modifications would provide some difficulty, yet with enough pre-screening of such modifications, the system should still prove accurate. Also, impression scans that are taken in the field can be stored and tagged with a vehicles's license plate number, and future inspections of that vehicle can be compared to already screened and cleared impressions of the same vehicle in order to search for variance.

Modern unmanned vehicles (UV) are complex systems. The current generation of UVs have extensive capabilities allowing the UV to sense its environment, create an internal representation of the environment, navigate within this environment by commanding movement and accomplish this in real-time. This proliferation of UV capabilities has resulted in large and complex software systems that are often distributed across multiple processors. Such systems have a reputation for convoluted implementations that result in software that is difficult to understand, expand, debug and repair. In order for a UV to operate successfully this issue of complex distributed software systems must be mastered. The computing science field views a modular, component based design as the best approach for implementing complex distributed software systems. Methodologies and toolkits such as frameworks and middleware have been developed to enable and simplify the implementation of distributed software systems. DRDC and other research institutions are developing UVs frameworks using CORBA middleware. A CORBA interface enables location transparency, thus it does not matter whether the component is locally or remotely located. The UV autonomy framework developed at DRDC is based upon the Miro framework which was originally developed for soccer playing robots. The Miro framework implements many key features and methods required by autonomous systems and Miro's basis in CORBA allows it to be easily modified and extended to support the unique requirements of military UVs. DRDC has modified and extended Miro so that it now supports autonomous unmanned ground vehicles. The process of implementing these changes substantiated the advantages of frameworks and middleware since Miro proved to be highly flexible and easy to extend.

Mobile robot designers frequently look to computer vision to solve navigation, obstacle avoidance, and object detection problems such as those encountered in parking lot surveillance. Stereo reconstruction is a useful technique in this domain and can be done in two ways. The first requires a fixed stereo camera rig to provide two side-by-side images; the second uses a single camera in motion to provide the images. While stereo rigs can be accurately calibrated in advance, they rely on a fixed baseline distance between the two cameras. The advantage of a single-camera method is the flexibility to change the baseline distance to best match each scenario. This directly increases the robustness of the stereo algorithm and increases the effective range of the system. The challenge comes from accurately rectifying the images into an ideal stereo pair. Structure from motion (SFM) can be used to compute the camera motion between the two images, but its accuracy is limited and small errors can cause rectified images to be misaligned. We present a single-camera stereo system that incorporates a Levenberg-Marquardt minimization of rectification parameters to bring the rectified images into alignment.

In this paper, a new real-time and intelligent mobile robot system for path planning and navigation using stereo camera embedded on the pan/tilt system is proposed. In the proposed system, face area of a moving person is detected from a sequence of the stereo image pairs by using the YCbCr color model and using the disparity map obtained from the left and right images captured by the pan/tilt-controlled stereo camera system and depth information can be detected. And then, the distance between the mobile robot system and the face of the moving person can be calculated from the detected depth information. Accordingly, based-on the analysis of these data, three-dimensional objects can be detected. Finally, by using these detected data, 2-D spatial map for a visually guided robot that can plan paths, navigate surrounding objects and explore an indoor environment is constructed. From some experiments on target tracking with 480 frames of the sequential stereo images, it is analyzed that error ratio between the calculated and measured values of the relative position is found to be very low value of 1.4 % on average. Also, the proposed target tracking system has achieved a high speed of 0.04 sec/frame for target detection and 0.06 sec/frame for target tracking.

For unmanned combat forces, some research proposes multi-robot coordination through common analytical coordination algorithms using reliable, high bandwidth communications. Such coordination is capable of optimal or near-optimal distribution of unmanned forces, but requires reliable communications and frequent feedback and control to ensure predictable performance. Others propose local autonomy, reducing dependence on reliable communications through greater intelligence within each unmanned system, but at the cost of optimality, predictability, and dependency on rich, high rate sensing. Thus a fundamental problem of swarm control can be starkly drawn. If centralized control is not practical and the swarm must function at a level comparable to manned forces, swarm members must adhere to common goal direction semantics that permits each unit to dissect its contribution to the team objective with or without consultation and negotiation. How, then should missions be expressed, allocated, and monitored?

Packaging is used to protect and enable intelligent sensor systems utilized in manned/unmanned ground vehicle systems/subsystems. Because Micro electro mechanical systems (MEMS) are used often in these sensor or actuation products, it must interact with the surrounding environment, which may be in direct conflict with the desire to isolate the electronics for improved reliability/durability performance. For some very simple devices, performance requirements may allow a high degree of isolation from the environment (e.g., stints and accelerometers). Other more complex devices (i.e. chemical and biological analysis systems, particularly in vivo systems) present extremely complex packaging requirements. Power and communications to MEMS device arrays are also extremely problematic. The following describes the research being performed at the U.S. Army Research, Development, and Engineering Command (RDECOM) Tank and Automotive Research, Development, and Engineering Center (TARDEC), in collaboration with Wayne State University, in Detroit, MI. The focus of the packaging research is limited to six main categories: a) provision for feed-through for electrical, optical, thermal, and fluidic interfaces; b) environmental management including atmosphere, hermiticity, and temperature; c) control of stress and mechanical durability; d) management of thermal properties to minimize absorption and/or emission; e) durability and structural integrity; and f) management of RF/magnetic/electrical and optical interference and/or radiation properties and exposure.

In this paper, the application of wheeled mobile robot (WMR) formation control in diffusion process characterization and control is discussed. We present a review over the current approaches on mobile robot formation control. A new consideration is presented on formation control within the framework of networked control system with wireless communication. The potential benefits of robot formation in distributed diffusion process measurement and control are discussed. In this paper, we present a new nonlinear control law for a general formation that can be useful in diffusion process boundary measurement. Then, we introduce our on-going project called Mobile Actuator and Sensor Network (MAS-net) on the diffusion process characterization and control. Experiment results are presented to illustrate how pattern formation can be achieved in MAS-net.

Recent experiences with robots in Iraq have proven that robotic technology is useful to the warfighter, but tools are needed to rapidly respond to evolving missions. This paper details a methodology for generation of software wrappers using JAUS to simplify prototyping and development of robotic systems (distributed, embedded and real-time system software modules). Software wrappers will allow insertion of modules into a visual prototyping environment. The wrappers will intercept module functions and bind them with functions needed to exercise the modules outside of the native environment. Generation of JAUS wrappers will enhance the development environment by reducing rote work and producing consistently behaving module interfaces. The resulting methodology will provide a rapid prototyping environment for use in sensor integration, Operator Control Unit (OCU) development and autonomous vehicle control.

Under the DARPA MARS 2020 program, Perceptek has developed a technical foundation for performing roadway operations in both structured and unstructured environments. Fully autonomous roadway operations require a large set of atomic functionalities that must seamlessly perform in concert in complex and dynamic environments. PercepTek has developed multiple atomic functionalities and implemented a robot control architecture called ARTEA (Autonomous Robotic Test and Evaluation Architecture) that blends these atomic functionalities into one cohesive system that can perform complex missions and reason about its environment. Some of the atomic functionalities that have been implemented and integrated onto our robotic test platform are vision-based road-following for both structured and unstructured roads, vision/radar-based vehicle-following, safety gap maintenance, road feature detection and response, road sign detection/ recognition, and pedestrian detection. In this paper, technical details of each of the individual robotic functionalities are presented along with their performance and limitations. We then discuss some of the critical components of the ARTEA architecture that are used for blending inputs from disparate functionalities and performing reasoning about the environment. Field testing was a critical aspect of our development process and we will discuss the test platform that was used to develop and test our robotic system. At the culmination of our MARS 2020 effort we performed a robotic test drive from Denver Colorado to New Orleans Louisiana in which we tested and evaluated various aspects of our system. We finally discuss our performance and the limitations of our system for this drive.

We propose the use of virtual wheels as the starting point of a new vehicle design. Each virtual wheel incorporates a pair of simple legs that, by simulating the rotary motion and ground contact of a traditional wheel, combine many of the benefits of legged and wheeled motion. We describe the use of virtual wheels in the design of a robotic mule, presenting an analysis of the mule's mobility the results of our efforts to model and build such a device.

Foster-Miller's unmanned ground vehicle, TALON, was originally developed under DARPA's Tactical Mobile Robotics (TMR) program. TALON has evolved over the years and has proven to be a robust, mobile, universal platform. As a result of the advances made in the evolution of TALON, new and far-reaching opportunities have been realized for unmanned ground vehicles. In recent conflicts such as in Afghanistan and Iraq, unmanned systems have played an important role and have extended the reach and capabilities of the War fighter. Technological advances have transformed unmanned vehicles in to useful tools and in some cases are used in lieu of sending in a soldier. Unmanned ground vehicles have seen recent and persistent success, as shown in theater, in the explosive ordinance disposal (EOD) and improvised ordinance disposal (IED) missions. Foster-Miller's TALON has experienced over ten thousand EOD and IED missions in Iraq alone. The success of the unmanned system has resulted in the doctrine "Send the robot in first". Foster-Miller has taken the role of the unmanned vehicle in yet another direction. Foster-Miller has transformed the TALON from a "practical" to "tactical" system. Through the combined efforts of Foster-Miller and the US Army, TALON has been involved in a weaponization program. To date, Foster-Miller has outfitted the TALON with 11 systems. As one can see, the unmanned ground vehicle is much more than a mobility platform.

The PackBot EOD firing system was developed to allow the safe initiation of disruptors used to disable improvised explosive devices (IEDs). Safety is assured despite the use of communication links shared by possibly insecure computers on UGVs and control stations. This specific firing system was developed to enforce a particular EOD operational protocol where a single operator is responsible for the loading and use of the weapon. A unique combination of hardware and software design allows for unprecedented safety.

Typical obstacles are walls, rocks, water plants and other nearby robots for a group of small scale fish robots and submersibles that have been constructed in our lab. Sonar sensors are not employed to make the robot structure simple enough. All of circuits, sensors and processor cards are contained in a box of 9 x 7 x 4 cm dimension except motors, fins and external covers. Therefore, image processing results are applied to avoid collisions. However, it is useful only when the obstacles are located far enough to give images processing time for detecting them. Otherwise, acceleration sensors are used to detect collision immediately after it happens. Two of 2-axes acceleration sensors are employed to measure the three components of collision angles, collision magnitudes, and the angles of robot propulsion. These data are integrated to calculate the amount of propulsion direction change. The angle of a collision incident upon an obstacle is the fundamental value to obtain a direction change needed to design a following path. But there is a significant amount of noise due to a caudal fin motor. Because caudal fin provides the main propulsion for a fish robot, there is a periodic swinging noise at the head of a robot. This noise provides a random acceleration effect on the measured acceleration data at the collision. We propose an algorithm which shows that the MEMS-type accelerometers are very effective to provide information for direction changes in spite of the intrinsic noise after the small scale fish robots have made obstacle collision.