The US Army Future Combat System (FCS) will implement Unmanned Ground Vehicles (UGV) in numbers not previously seen before in military operations. Many of these vehicles will also be larger and faster than the small robots typically used today for explosive ordnance disposal and general improvised explosive device handling. More importantly, FCS will implement these UGV's in scenarios were they will be in much closer proximity to soldiers and other non-combatant personnel. This paper describes the plan for developing an appropriate match of technology for autonomous UGV maneuver with the emerging need for safety release verification for these systems prior to fielding. The plan is followed by descriptions of initial data collections with a UGV, that will form the starting point in this safety release process, and stimulate further use and refinement of this process for large UGV's in applications beyond FCS.

Due to the technology available, most previous work in teleoperated robotics used relatively low-resolution video links and provided limited perceptual feedback to the teleoperator. In most cases, these projects reported only limited teleoperator success compared to vehicles with human drivers on-board. We set out to build a high-fidelity teleoperation system which takes advantage of recent technological advances. This system permits highly capable teleoperation and has allowed us to begin to investigate the minimum system requirements for effective teleoperation.

The Black Knight is a 12-ton, C-130 deployable Unmanned Ground Combat Vehicle (UGCV). It was developed to demonstrate how unmanned vehicles can be integrated into a mechanized military force to increase combat capability while protecting Soldiers in a full spectrum of battlefield scenarios. The Black Knight is used in military operational tests that allow Soldiers to develop the necessary techniques, tactics, and procedures to operate a large unmanned vehicle within a mechanized military force. It can be safely controlled by Soldiers from inside a manned fighting vehicle, such as the Bradley Fighting Vehicle. Black Knight control modes include path tracking, guarded teleoperation, and fully autonomous movement. Its state-of-the-art Autonomous Navigation Module (ANM) includes terrain-mapping sensors for route planning, terrain classification, and obstacle avoidance. In guarded teleoperation mode, the ANM data, together with automotive dials and gages, are used to generate video overlays that assist the operator for both day and night driving performance. Remote operation of various sensors also allows Soldiers to perform effective target location and tracking. This document covers Black Knight's system architecture and includes implementation overviews of the various operation modes. We conclude with lessons learned and development goals for the Black Knight UGCV.

We present the Crusher system for autonomously navigating complex off-road terrain. In this paper, we de-scribe the Crusher system’s three-pronged approach for safely and reliably moving between widely-spaced waypoints. First, the system automatically interprets aerial map data to assess mobility risk and plans trajectories that move from way point to way point. Second, the system uses a ladar- and camera-based perception system to detect and avoid hazards that are not discernable in the map data or that appear after the area is mapped. Third, the Crusher vehicle breaches hazards too difficult for on-board sensors to detect. The autonomy software is adaptive for operation on a wide range of terrain types. To date, the Crusher system has been tested on terrain with dense trees, long washes, deep ditches, steep slopes, thick brush, and large rocks. In early 2007, the system was tested at Ft. Bliss in Texas, where it drove over two hundred fifty kilometers autonomously.

Radar offers advantages as a robotic perception modality because it is not as vulnerable to the vacuum, dust, fog, rain, snow and light conditions found in construction, mining, agricultural and planetary-exploration environments. However radar has shortcomings such as a large footprint, sidelobes, specularity effects and limited range resolution—all of which result in poor environment maps. The fusion of successive radar observations can alleviate radar shortcomings and improve map fidelity. Sensor models exist for the fusion of sonar, laser and stereo into evidence grids. However radar outputs richer data and cannot use those existing models. This paper presents a sensor model that uses constant false alarm for occupancy detection and incorporates heuristic rules to approach occlusions. The resulting radar-based map of outdoors can suit robot obstacle avoidance, navigation and tool deployment. The limited map detail achieved suggests the need for a more rigorous probabilistic approach to encode the dependencies and estimate the model parameters.

The National Robotics Engineering Consortium and UltraStrip Systems Inc. have developed a highly flexible and productive robot to strip paint from large ships and other large ferro-magnetic structures based on the patents obtained by UltraStrip Systems, Inc. (US patents: 6,425,340; 5,849,099; 5,628,271). Removal of corrosion and coatings from large vessels has become a serious economic and environmental problem, and current practices are becoming infeasible. The M2000 robot removes paint from ships using ultrahigh pressure water jets and recovers the water and debris in an environmentally sound way. The addition of simple, easy-to-use, cruise control features to the robot has permitted significant increases in productivity, safety, and stripping quality.

Material Classification By Drilling Diana LaBelle, John Bares, Illah Nourbakhsh Pages 1-6 (2000 Proceedings of the 17th ISARC, Taipei, Taiwan, ISBN 9789570266986, ISSN 2413-5844) Abstract: Underground coal mining is one of the most dangerous occupations. Years of effort have been dedicated to researching methods of characterizing mine roof and floor for improving the mining environment. This research investigates using a neural network to classify rock strata based on the physical parameters of a roof bolting drill. This paper presents our methodology, as well as early results based on drilling experiments conducted in the laboratory using a custom poured concrete test block. We have classified, with a trained network, the five layers of the test block with less than 5% error. Keywords: mining automation, rock classification, neural networks, drilling, coal interface detection DOI: https://doi.org/10.22260/ISARC2000/0088 Download fulltext Download BibTex Download Endnote (RIS) TeX Import to Mendeley

This special issue contains representative works from the Second International Conference on Field and Service Robotics that was held in Pittsburgh, Pennsylvania on August 29-31, 1999. This conference convened researchers and end-users to discuss issues facing the transition of field and service robotics from laboratory to daily use. These papers emphasize both contributions at the component level as well as system performance in target applications. A requirement for paper selection was demonstrated field results. We believe that the papers in this special issue represent the state of the art in field and service robotics. S. Thrun et al.’s recent demonstrations of their museum tour guide robot has raised the bar on mobile robot performance in cluttered and dynamic environments. The key technology addressed is that of robot positioning using on-board range sensors. Their approach pushes the capabilities of probabilistic algorithms and demonstrates the results in a dynamic museum environment. D. Yoerger and his collaborators have collected impressive data from their underwater robot in an ongoing quest to map and understand subsea environments. One of the many critical challenges addressed in Yoerger’s work is image mosaicing when vehicle motion is not well known. Just like Thrun in the case of indoor mobile robotics, Yoerger sets the standard to which underwater robot performance will be compared. Back on terra firma, D. Apostolopoulos et al.’s Nomad robot operated under extremely harsh conditions during its search for meteorites in Antarctica. While contending with the standard challenges of positioning and navigation, they also consider probabilistic techniques for identifying meteorites. (False positives are appropriately termed meteorwrongs!) The Nomad robot demonstration in Antarctica is one step toward the ultimate goal of sending a mobile robot to the moon that can be supervised from earth. Infrastructure maintenance is a huge and growing market: Kirkham et al. developed the PIRAT in-pipe robot system for the purpose of sewer inspection with minimal invasion. Their approach to inspection is highly cost-effective and provides a high-resolution laser range image of the pipe that enables accurate condition assessment. Due to the difficulty of direct human interpretation of the huge amount of data, the authors have implemented an automated scheme to filter the data and detect problematic locations in the pipeline. Excellent results are described for several types of sewers. Another brand of robotics research is aimed at directly helping people in their daily lives. G. Lacey and S. MacNamara describe PAM-AID, a robotic smart walker to assist the elderly. Results from a series of nursing home tests indicate that this kind of aid can be accepted and can help aging and partially disabled people attain an improved quality of life. This special issue also addresses the state of the art in mobile robot component technologies for motion, vision, and positioning. Most mobile mechanisms are nonholonomic. For example, a differential drive robot is nonholonomic because it cannot control velocities perpendicular to its heading. The mobile mechanism that R. Holmberg and O. Khatib introduce does not have such a nonholonomic constraint and therefore allows for a richer set of motions that the robot can experience, which is especially useful in highly cluttered environments. Another key technology is D. Langer et al.’s three-dimensional scanning laser radar. This system is primarily intended for high-precision mapping applications and results are shown for a variety of applications including mapping of tunnels and manufacturing work cells. A unique dual frequency scheme enables extremely high resolution as well as long range, low noise, and low error returns for a wide range of target materials and surface conditions. Even when robots can see about themselves, they commonly require high speed means to measure position information. S. Sukkarieh et al. describe a low-cost, high-precision, strap down inertial unit that is designed with the specific goal of improving fault rejection and isolation. Another approach to vehicle positioning that relies on camera vision is described by A. Kelly. He has developed a practical, high-performance mobile robot localization technique that exploits the fact that many manmade environments are composed of substantially flat, visually textured surfaces. The mobile robot drives on these surfaces and, using vision and mosaicing approaches, can localize itself without using inertial, GPS, or other external positioning devices. Kelly demonstrates the technique in an industrial material handling setting. S. Hirose, perhaps more than any other robotics researcher, has shown that a mobile robot is more than a garbage-can-shaped autonomous device. Hirose’s mobile robots look like members of a “robot zoo” from your favorite science fiction movie. Some of his robots are legged, some are wheeled, and some are both! In the early 1970s, Hirose introduced a radical alternative to wheeled and legged mechanism with the first snake robot ever built. His paper in this journal puts forth a design methodology, termed variable constraint mechanism, that he has applied to a number of recent robots to extend their capability in widely varying terrain and conditions. We hope that you enjoy and benefit from this set of excellent papers as much as we have!

Excavators are used for the rapid removal of soil and other materials in mines, quarries, and construction sites. The automation of these machines offers promise for increasing productivity and improving safety. To date, most research in this area has focussed on selected parts of the problem. In this paper, we present a system that completely automates the truck loading task. The excavator uses two scanning laser rangefinders to recognize and localize the truck, measure the soil face, and detect obstacles. The excavator's software decides where to dig in the soil, where to dump in the truck, and how to quickly move between these points while detecting and stopping for obstacles. The system was fully implemented and was demonstrated to load trucks as fast as human operators.

Dante II is a unique walking robot that provides important insight into high-mobility robotic locomotion and remote robotic exploration. Dante II’s uniqueness stems from its combined legged and rappelling mobility system, its scanning-laser rangefinder, and its multilevel control scheme. In 1994 Dante II was deployed and successfully tested in a remote Alaskan volcano, as a demonstration of the fieldworthiness of these technologies. For more than five days the robot explored alone in the volcano crater using a combination of supervised autonomous control and teleoperated control. Human operators were located 120 km distant during the mission. This article first describes in detail the robot, support systems, control techniques, and user interfaces. We then describe results from the battery of field tests leading up to and including the volcanic mission. Finally, we put forth important lessons which comprise the legacy of this project. We show that framewalkers are appropriate for rappelling in severe terrain, though tether systems have limitations. We also discuss the importance of future “autonomous” systems to realize when they require human support rather than relying on humans for constant oversight.

We present Darwin2K, a widely-applicable, extensible software tool for synthesizing and optimizing robot configurations. The system uses an evolutionary optimization algorithm that is independent of task, metrics, and type of robot, enabling the system to address a wide range of synthesis problems. Darwin2K can synthesize fixed-base and mobile robots (including free-flying robots, mobile manipulators, modular robots, and multiple or bifurcated manipulators), and includes a toolkit of simulation and analysis algorithms which are useful for many synthesis tasks. Some of these capabilities, such as dynamic simulation, are novel in automated synthesis of robots. An extensible software architecture enables new synthesis tasks to be addressed while maximizing use of existing system capabilities; this extensibility is a key contribution of the system. A key challenge is effectively optimizing multiple performance metrics; we present a method called Requirement Prioritization that guides the evolutionary algorithm through the design space. We apply Darwin2K to a robot synthesis tasks that includes synthesis of robot kinematics, dynamics, structural geometry, and actuator selection to meet and optimize multiple performance requirements. Introduction: Why Automated Synthesis? Robot configuration design is characterized by generating an artifact which is capable of motion in an area or volume. Typically, the configuration process generates the overall form of the robot, including kinematics and other geometry at the bare minimum but often including approximate descriptions of inertial properties, actuator and material selection, and structural geometry. This design process is often performed in an ad hoc manner. It can be difficult to translate the requirements of a task into a robot configuration, and in contrast to some other engineering disciplines there are few design rules that can simplify this process. Human designers rely on intuition and experience with related design problems, and on engineering rules based on the experience of other expert designers. When designing a robot for a task with many new characteristics, relevant experience may be quite limited and may unnecessarily restrict the range of designs that are explored. Frequently, a human investigates a small number of concepts on paper and selects a few that look promising. More detailed studies may then be performed on these, culminating in the simulation of one or more designs. One of the candidates is selected for detailed design, with tools such as finite element analysis used to evaluate parts of the design. Once the robot is built, changes may be required due to unforeseen problems or interactions s that were not modeled in simulation. Significant design iterations are often not practical, since much of a project’s schedule and resources may be devoted to creating a single robot; building a second or third robot to remedy design flaws is out of the question for many robot design projects. Thus, it is crucial to perform as much analysis and simulation as possible before the robot is built, and it is highly desirable to “get it right” the first time -since the first time may be the only time. Because of these factors, automated synthesis tools are especially attractive for robot design. Synthesis tools can address design problems for which there are little or no relevant human experience, can explore much larger numbers and ranges of designs than a human, can quickly perform design iterations in simulation, and can produce a well-optimized solution with high confidence of performance, all of which contribute to the likelihood of success of the first physical implementation. Darwin2K is a software toolkit for automated synthesis of robot configurations. It includes capabilities for quickly describing and modifying robot configurations, simulating configurations as they perform tasks, and automatically synthesizing configurations to meet task-specific requirements and to optimize performance. Darwin2K is very useful in the early stages of the configuration process, as it can automatically explore tens of thousands of designs and can allow a human designer to rapidly perform design iterations and evaluate potential robots in simulation.

Robot configuration design is hampered by the lack of established, well-knowndesign rules, and designers cannot easily grasp the space of possible designs and the impactof all design variables on a robot's performance. Realistically, a human can only designand evaluate several candidate configurations, though there may be thousands ofcompetitive designs that should be investigated. In contrast, an automated approach toconfiguration synthesis can create tens of thousands of designs and...

Hydraulic machines used in mining and excavation applications are nonlinear systems. Apart from the nonlinearity due to the dynamic coupling between the different links there are significant actuator nonlinearities due to the inherent properties of the hydraulic system. Optimal motion planning for these machines, i.e., planning motions that optimize a user-selectable combination of criteria such as time. energy, etc., would help the designers of such machines, besides aiding the development of more productive robotic machines. Optimal motion planning in turn requires fast (computationally efficient) machine models in order to be practically usable. This work proposes a method for constructing hydraulic machine models using memory-based learning. We demonstrate the approach by constructing a machine model of a 25-ton hydraulic excavator with a 10 m maximum reach. The learning method is used to construct the hydraulic actuator model and is used in conjunction with a linkage dynamic model to construct a complete excavator model that is much faster than an analytical model. Our test results show an average bucket tip position prediction error of 1 m over 50 sec of machine operation. This is better than any comparable speed model reported in the literature. The results also show that the approach effectively captures the interactions between the different hydraulic actuators. The excavator model is used in a time-optimal motion planning scheme. We demonstrate the optimization results on a real excavator testbed to underscore the effectiveness of the model for optimal motion computation.

Underground coal mining is an industry well suited for robotic automation. Human operators are severely hampered in dark, dusty, and cramped mines, and productivity suffers. Even a slight improvement in productivity can amount to thousands of dollars of additional revenue per machine per day. Automation to date has relied on infrastructure to guide the equipment. The industry finds this approach unsuitable, and it has not taken root. Our approach uses machine-mounted video cameras to guide the equipment. It utilizes natural infrastructure and equipment com- monly used in mines. We have demonstrated that our approach meets the requirements for cutting straight entries and mining the proper amount of coal per cycle. The technology is rapidly approaching beta form and will be deployed in several mines in the coming months.