This paper describes the stereo vision near-field terrain mapping system used by the Legged Squad Support System (LS3) quadruped vehicle to automatically adjust its gait in complex natural terrain. The mapping system achieves high robustness with a combination of stereo model-based outlier rejection and spatial and temporal filtering, enabled by a unique hybrid 2D/3D data structure. Classification of sparse structures allows the vehicle to traverse through vegetation. Inference of negative obstacles allows the vehicle to avoid steep drop-offs. A custom designed near-infrared illumination system enables operation at night. The mapping system has been tested extensively with controlled experiments and 72km of field testing in a wide variety of terrains and conditions.

This paper presents an approach to accurate and scalable multiple-model (MM) state estimation for hybrid systems with intermittent, cyclical, multimodal dynamics. The approach consists of using discrete-state estimation to identify a system's dynamical and behavioral contexts and determine which motion models appropriately represent current dynamics and which individual and MM filters are appropriate for state estimation. Furthermore, the heirarchical structure of the dynamics is explicitly encoded, which enables detection not only of rapid transitions between motion models but of higher level behavioral transitions as well. This improves the accuracy and scalability of conventional MM state estimation, which is demonstrated experimentally on a mobile robot that exhibits fast-switching, multimodal dynamics.

BigDog is a four legged robot with exceptional rough-terrain mobility. In this paper, we equip BigDog with a laser scanner, stereo vision system, and perception and navigation algorithms. Using these sensors and algorithms, BigDog performs autonomous navigation to goal positions in unstructured forest environments. The robot perceives obstacles, such as trees, boulders, and ground features, and steers to avoid them on its way to the goal. We describe the hardware and software implementation of the navigation system and summarize performance. During field tests in unstructured wooded terrain, BigDog reached its goal position 23 of 26 runs and traveled over 130 meters at a time without operator involvement.

Running is a complex dynamic task that places strict requirements on both the physical components and software-control systems of a robot. This paper explores some of those requirements and, in particular, explores how a variable-compliance actuation system can satisfy many of them. We present the mechanical design and software-control system for such an actuator system. We analyze its performance through simulation and bench-top experimental validation of a prototype version. In conclusion, we demonstrate, through simulation, the application of our proof-of-concept actuator to the problem of biped running.

This paper describes the development of a legged robot designed for general locomotion of complex terrain but specialized for dynamical, high-speed climbing of a uniformly convex cylindrical structure, such as an outdoor telephone pole. This robot, the RiSE V3 climbing machine-mass 5.4 kg, length 70 cm, excluding a 28 cm tail appendage-includes several novel mechanical features, including novel linkage designs for its legs and a non-backdrivable, energy-dense power transmission to enable high-speed climbing. We summarize the robot's design and document a climbing behavior that achieves rapid ascent of a wooden telephone pole at 21 cm/s, a speed previously unachieved-and, we believe, heretofore impossible-with a robot of this scale. The behavioral gait of the robot employs the mechanical design to propel the body forward while passively maintaining yaw, pitch, and roll stability during climbing locomotion. The robot's general-purpose legged design coupled with its specialized ability to quickly gain elevation and park at a vertical station silently with minimal energy consumption suggest potential applications including search and surveillance operations as well as ad hoc networking.

Complete automation of trajectory planning tools for material deposition/removal applications has become increasingly necessary to reduce the ldquoconcept-to-consumerrdquo timeline for rapid product introduction in industries such as the automotive industry. The work in this paper is specifically motivated by automotive spray painting. Prior developments in automated trajectory planning tools promise to reduce the time required to program the robots; however, these approaches are limited to surfaces that are either approximately planar or topologically simple (i.e., with no holes). To extend the applicability of these planning tools to nonplanar and topologically complex surfaces, currently the user has to manually segment a complex surface into simpler subsets, i.e., subsets that are approximately extruded surfaces and contain no holes. However, the complex nature of the relationships between surface segmentation and resulting output characteristics such as material deposition uniformity, process cycle time, and material waste makes the task of manually segmenting the surface difficult. In this paper, we develop a hierarchical procedure that automatically segments a surface based on surface geometry, surface topology, and path geometry to obtain topologically simple subsets that are approximately extruded surfaces. Finally, we compare the effectiveness of our segmentation with the state of the art on a few automotive surfaces in simulation.

This paper addresses the coupled navigation and control problem for fully actuated dynamical systems operating in cluttered environments. The proposed method composes local feedback control policies to address the global problem in a formally correct manner. Our contribution extends the method of sequential composition to a new class offlow-throughpolicies, which are defined for an idealized point robot that is subject to bounds on velocity and acceleration. These policies, which are defined over convex polytopes, express natural behaviors and guarantee safe trajectories while respecting system constraints. This paper presents a constructive technique for building hybrid control systems with guaranteed transitions and stability properties. Simulation results demonstrate policy-based decision making for various initial conditions.

This paper discusses the design principles and philosophy of the BiMASC, a biped with mechanically adjustable series compliance which incorporates tuned mechanical leg springs. This robot will be capable of dynamic running using mechanical leg springs, as well as dynamic ballistic walking with human-like passive leg swing behavior. The BiMASC will enable the study of the role of both controllable compliance in running and will serve as a test platform for control strategies that utilize the leg springs and other natural dynamics of the robot. The mechanism is designed to behave in a dynamically "clean" manner, such that relatively simple mathematical models will accurately predict the robot's behavior. The availability of simple and accurate mathematical models will facilitate the design of controllers, accurate simulations, and the implementation of accurate model-based control on the robot.

Outside of the laboratory, accurate models of ground impact dynamics are either difficult or impossible to obtain. Instead, a rigid ground model is often used in gait and controller design, which simplifies the system model and allows attention to remain focused on other aspects of running. In real-world terrain this simplification may overlook important dynamic effects. Immediately following a foot touchdown event, sensitivity to ground stiffness is at its highest and at the same time the accuracies of state estimates are at their lowest. Even if ground stiffness is known and state estimates are accurate, actuator bandwidth limitations make immediate compensation difficult. Taking inspiration from nature, we propose a novel solution to attenuate the effects of unexpected ground stiffness changes using a unified control system comprised of hardware passive dynamics and open-loop software control policies.

What would it be like if we could give our robot high level commands and it would automatically execute them in a verifiably correct fashion in dynamically changing environments? This work demonstrates a method for generating continuous feedback control inputs that satisfy high-level specifications. Using a collection of continuous local feedback control policies in concert with a synthesized discrete automaton, this paper demonstrates the approach on an Ackermann-steered vehicle that satisfies the command "drive around until you find an empty parking space, then park." The system reacts to changing environmental conditions using only local information, while guaranteeing the correct high level behavior. The local policies consider the vehicle body shape as well as bounds on drive and steering velocities. The discrete automaton that invokes the local policies guarantees executions that satisfy the high-level specification based only on information about the current availability of the nearest parking space. This paper also demonstrates coordination of two vehicles using the approach.

This paper presents an approach to accurate and scalable multiple-model state estimation for hybrid systems with intermittent, multi-modal dynamics. The approach consists of using discrete-state estimation to identify a system’s behavioral context and determine which motion models appropriately represent current dynamics, and which multiple-model filters are appropriate for state estimation. This improves the accuracy and scalability of conventional multiple-model state estimation. This approach is validated experimentally on a mobile robot that exhibits multi-modal dynamics.

This paper introduces the concept of gait transitions, acyclic feedforward motion patterns that allow a robot to switch from one gait to another. Legged robots often utilize collections of gait patterns to locomote over a variety of surfaces. Each feedforward gait is generally tuned for a specific surface and set of operating conditions. To enable locomotion across a changing surface, a robot must be able to stably change between gaits while continuing to locomote. By understanding the fundamentals of gaits, we present methods to correctly transition between differing gaits. On two separate robotic platforms, we show how the application of gait transitions enhances each robot's behavioral suite. Using the RHex robotic hexapod, gait transitions are used to smoothly switch from a tripod walking gait to a metachronal wave gait used to climb stairs. We also introduce the RiSE platform, a hexapod robot capable of vertical climbing, and discuss how gait transitions play an important role in achieving vertical mobility

In this paper we generate gaits for dynamics systems that are subject to non-holonomic velocity constraints. These systems are referred to as mixed non-holonomic systems. The motion of such systems is governed by both the non-holonomic constraints acting on the system and a system of differential equations constraining the evolution of generalized momentum. We propose a method that utilizes both governing motions, that is, satisfying all the constraints and instantaneously conserving momentum along un-restricted directions, to generate gaits for systems like the snakeboard, which belongs to the family of mixed non-holonomic systems. We accomplish this by defining a new scaled momentum variable. This scaled momentum allows us to easily explore the design of gaits that causes momentum to evolve such that a desired non-trivial motion results

Motivated by an error-recovery locomotion problem, we propose a control technique for a complex mechanical system by decomposing the system dynamics into a collection of simplified models. The robot considered, The Rocking and Rolling Robot (RRRobot), is a high-centered round-bodied robot that locomotes on a plane by swinging its legs and rocking on its shell. We identify the elements contributing to locomotion through two steps: 1) decoupling the leg-body rotation dynamics from the body-plane contact kinematics, and 2) decoupling the body rotational dynamics into dynamics along each rotational axis. We show, using simulation, that such decoupling provides a good approximation to RRRobot's locomotion and use these models to find an approximate control solution for RRRobot: a mapping between planar translation and leg motions

Complete automation of trajectory planning tools for material deposition/removal applications has become increasingly necessary to reduce the “concept-to-consumer” timeline in product development. Segmentation of a complex automotive surfaces into topologically simple surfaces remains a barrier in the automation of trajectory generation. In this paper, we develop a novel hierarchical procedure to segment a complex automotive surface into geometrically as well as topologically simple components.

Constructing climbing behaviors for hyper-redundant robots that account for the system dynamics requires a model of robot dynamics under contact and friction. One common model, rigid-body dynamics with coulomb friction, unfortunately is both an ambiguous and inconsistent set of dynamic axioms. This paper addresses the ambiguity problem by developing an algorithm which computes the set of joint torques such that all solutions are guaranteed to produce the desired system behavior. This algorithm is applied to a type of robot climbing which we denote climbing via bracing where a hyper-redundant robot stabilizes itself against gravity by pressing outward to induce friction. By bracing with a fraction of the robot, which we term a brace the remainder of the robot remains free to move upward and brace at a higher location. A sequence of braces thus moves the robot upward.

This paper proposes an approach to robust state estimation for mobile robots with intermittent dynamics. The approach consists of identifying the robot’s mode of operation by classifying the output of onboard sensors into mode-specific contexts. The underlying technique seeks to efficiently use available sensor information to enable accurate, high-bandwidth mode identification. Context classification is combined with multiple-model filtering in order to significantly improve the accuracy of state estimates for hybrid systems. This approach is validated in simulation and shown experimentally to produce accurate estimates on a jogging robot using low-cost sensors.

This research is focused on developing trajectory planning tools for the automotive painting industry. The geometric complexity of automotive surfaces and the complexity of the spray patterns produced by modern paint atomizers combine to make this a challenging and interesting problem. This paper documents our efforts to develop computationally tractable analytic deposition models for electrostatic rotating bell (ESRB) atomizers, which have recently become widely used in the automotive painting industry. The models presented in this paper account for both the effects of surface curvature as well as the deposition pattern of ESRB atomizers in a computationally tractable form, enabling the development of automated trajectory generation tools. We present experimental results used to develop and validate the models, and verify the interaction between the deposition pattern, the atomizer trajectory, and the surface curvature. Limitations of the deposition model with respect to predictions of paint deposition on highly curved surfaces are discussed. Note to Practitioners-The empirical paint deposition models developed herein, which are fit to experimental data, offer a significant improvement over models that are typically used in industrial robot simulations. The improved simulation results come without the computational cost and complexity of finite element methods. The models could be incorporated, as is, into existing industrial simulation tools, provided the users are cognizant of the model limitations with respect to highly curved surfaces. Although the models are based on readily available information, incorporating the models into existing robot simulation software would likely require support from the software vendor.

We show through experiment and simulation that a high-centered round-bodied legged robot can locomote by generating out-of-phase motions of reaction masses attached to its legs. These leg motions create body attitude oscillations which, when coupled with the slip-free contact constraints, locomote the robot. By varying the mean position of the leg oscillations, the robot can move in different directions in the plane. We also present some simplified models, where body attitude dynamics and contact kinematics are decoupled, to explain this form of legless locomotion.

Running is a complex dynamical task which places strict design requirements on both the physical components and software control systems of a robot. This paper explores some of those requirements and illustrates how a variable compliance actuation system can satisfy them. We present the design, analysis, simulation, and benchtop experimental validation of such an actuator system. We demonstrate, through simulation, the application of our prototype actuator to the problem of biped running.

Motivated by finding locomotion primitives for a legged robot, we present controllability results and a technique for kinematic reduction for a variable inertia mechanical system. We demonstrate configuration controllability for the system under consideration and use the symmetry resulting from angular momentum conservation to develop a kinematic representation of the mechanical system. We also show through simulation how plans for the kinematic representation can be implemented on the full dynamical mechanical system. It is hoped that this technique leads us to a general procedure for solving the gait synthesis problem.

For automated spray painting robots, the choice of a start curve from which all subsequent paths will be determined, plays a critical role in ensuring uniform coverage of the target surface. In this paper, we ultimately propose a method to determine the start curve for two different procedures of coverage path construction. For the first procedure, we average the target surface normal to optimize the average error between desired and resultant deposition with the added goal of minimizing cycle time. For the second technique, we make recourse to the Gauss-Bonnet theorem to minimize the non-uniformity of paint deposition and cycle time for the entire coverage path.

We propose a locomotion technique for a legged robot that is high-centered, i.e., a robot stuck on a block with its legs dangling in air. By using its legs as reaction masses, the robot might be able to rock and roll on its stomach and incrementally move forward off the block, a form of legless locomotion using halteres. With locomotion of high-centered robots using body attitude oscillations as motivation, this paper focuses on studying the interplay between leg motions and body roll-pitch-yaw dynamics. We present results from simulation of two simplified models in which body motion is restricted to the roll and roll-yaw space respectively.

This paper develops a method of composing simple control policies, applicable over a limited region in a dynamical system's free space, such that the resulting composition completely solves the navigation and control problem for the given system operating in a constrained environment. The resulting control policy deployment induces a global control policy that brings the system to the goal, provided that there is a single connected component of the free space containing both the start and goal configurations. In this paper, control policies for both kinematic and simple dynamical systems are developed. This work assumes that the initial velocities are somewhat aligned with the desired velocity vector field. We conclude by offering an outline of an approach for accommodating arbitrary dynamical constraints and initial conditions.

This paper presents preliminary development of autonomous sensor-guided behaviors for a six-legged dynamical robot (RHex). The behaviors represent the exteroceptive closed-loop locomotion strategies for this system. Simple motion models for RHex are described and used to deploy controllers for inertial straight line locomotion and visual line following. Results are experimentally validated on the robot.

As part of an ongoing collaborative effort with the Ford Motor Company, our research aims to develop practical and efficient trajectory planning tools for automotive painting. This paper documents our efforts to develop analytic deposition models for electrostatic rotating bell (ESRB) atomizers, which have recently become widely used in the automotive painting industry. Conventional deposition models, used in earlier automatic trajectory planning tools, fail to capture the complexity of deposition patterns generated by ESRB atomizers. The models presented here take into account both the surface curvature and the deposition pattern of ESRB atomizers, enabling planning tools to optimize trajectories to meet several measures of quality, such as coating uniformity. In addition to the development of our models, we present experimental results used to evaluate our models, and verify the interaction between the deposition pattern, trajectory, and surface curvature.

We address the task of covering a closed orientable surface embedded in /spl Rfr//sup 3/ without any prior information about the surface. For applications such as paint deposition, the effector (the paint atomizer) does not explicitly cover the target surface, but instead covers an offset surface-a surface that is a fixed distance away from the target surface. Just as Canny and others use critical points to look for changes in connectivity of the free space to ensure completeness of their roadmap algorithms, we use critical points to identify changes in the connectivity of the offset surface to ensure full surface coverage. The main contribution of this work is a method to construct unknown offset surfaces using a procedure, also developed in this paper, to detect critical points.

Documents our efforts to develop a scalable low-cost and high-performance network infrastructure for a novel distributed automation system-minifactory. The resulting communication system (AAA-Net) relies on a suite of low-latency protocols carried by a commercial 100 Mb Fast Ethernet network. In addition to documenting the operation of this network system and its associated protocols we offer a brief comparison to the currently available commercial field network systems, and present experimental verification of the performance delivered by AAA-Net under typical and extreme operating conditions.

We describe a new roadmap, termed the rod-HGVG, for motion planning of a rod-shaped robot operating in a three-dimensional volume. This roadmap is defined in terms of work space distance enabling us to prescribe an incremental construction procedure. This allows the rod to explore its configuration space, R/sup 3//spl times/S/sup 2/, without ever explicitly constructing the configuration space. In fact, the rod robot need not know the work space ahead of time. We term the rod-HGVG a piecewise retract because it comprises many retracts. Homotopy theory asserts that there cannot be in general a one dimensional retract of noncontractable five-dimensional space. Instead, we define an exact cellular decomposition on R/sup 3//spl times/S/sup 2/ and a retract in each cell. Next, we "connect" the retracts of each cell forming a piecewise retract of the rod's configuration space.

We describe the mechanical and electronic design of a precision, 2-DOF robot manipulator. The manipulator is one of a wide variety of modular robotic agents in the minifactory, a rapidly deployable precision assembly system under development in our laboratory. The manipulator, in cooperation with another type of 2-DOF robot, termed a courier, can perform 4-DOF assembly operations emulating a SCARA robot. This arrangement provides increased precision, higher throughput, smaller footprint, and increased flexibility relative to the SCARA robot.

A distributed cooperative coverage algorithm DC/sub R/ is presented, which is derived from an earlier complete single-robot algorithm, CC/sub R/. DC/sub R/ executes independently on each robot in a team where the individual robots do not know the initial locations of their peers and applies to systems of robots operating in a rectilinear environment that use only intrinsic contact sensing to determine the boundaries of the environment. Due to the reactive nature of CC/sub R/, the natural extension to DC/sub R/ preserves the completeness properties of the single-robot algorithm, and the outline of a completeness proof of DC/sub R/ is also presented. DC/sub R/ has been implemented in simulation, and directions for future work are presented which will make the algorithm more suited to physical robot systems.

We document our initial efforts to instantiate visually-guided cooperative behaviors between robotic agents in the minifactory environment. The minifactory incorporates high-precision, 2-DOF robotic agents to perform micron-level precision 4-DOF assembly tasks. Here we utilize two minifactory agents to perform visual servoing. We present a detailed description of the control and communication systems used to coordinate the agents. To provide a suitable communications infrastructure, we describe the development of a new interagent communication system which uses low-latency protocols carried by a commercial 100 Mb Ethernet network. Finally, we conclude by presenting experimental results front our first coordinated multi-agent task, the visually guided positioning of a small medical device.

Exact cellular decompositions are structures that globally encode the topology of a robot's free space, while locally describing the free space geometry. These structures have been widely used for path planning between two points, but can be used for mapping and coverage of robot free spaces. In this paper, we define exact cellular decompositions where critical points of Morse functions indicate the location of cell boundaries. Morse functions are those whose critical points are non-degenerate. Between critical points, the structure of a space is effectively the same, so simple control strategies to achieve tasks, such as coverage, are feasible within each cell. In this paper, we derive a general framework for defining decompositions in terms of critical points and then give examples, each corresponding to a different task. All of the results in this paper are derived in an m-dimensional Euclidean space, but the examples depicted in the figures are 2D and 3D for ease of presentation.

The need for high precision robotic systems has increased in recent years. Examples include automated assembly of complex electromechanical products such as disk drives, cameras, and medical devices. In this paper we present an optical sensor which can sense motion with better than 0.15 /spl mu/m resolution while also offering the advantages of insusceptibility to ambient lights, high bandwidth, and a large dynamic range. The sensor is used for calibration and coordination tasks in a novel agile robotic system (minifactory) which employs cooperative 2-DOF robots. The sensor can also be used for testing, calibrating, and position sensing in many other high precision robotic applications.

An existing hybrid control strategy is extended and applied to the control of a novel courier robot based on planar (Sawyer) motor technology augmented with integral sensing. The controller is designed to accommodate actuator limits and the desired motion characteristics, eliminating the need for careful matching of trajectory parameters with controller gains. Controller extensions allow for the unusual actuator constraints of the planar motors and independent adjustment of the parameters for each axis. Simulation results are presented for several cases to demonstrate the diversity of situations within the capability of the controller.

Industrial Sawyer motor technology has existed for nearly three decades, traditionally being operated as open-loop positioners. Such systems can attain micron level motion resolution and open-loop repeatability of 10 /spl mu/m. This paper outlines a optical sensing technique for Sawyer motors which is capable of sub-micron resolution. Such a sensor can easily be integrated into a motor enabling closed-loop control and sub-micron repeatability.

The goal of the architecture for agile assembly (AAA) is to enable rapid deployment and reconfiguration of automated assembly systems through the use of cooperating, modular, robust, robotic agents. AAA agent programs must be completely distributed and specify cooperative precision behavior in a structured, well known environment. Thus, the structure of agent programs is carefully designed to allow packaging of all the information necessary for coordinated execution when downloaded to a physical agent. To make the specification and execution of the potentially complex and fragile cooperative behaviors robust, our programs define ordered sets of control strategies and allow a low level real-time hybrid control system to sequence the strategies rather than burdening the agent program with the management of this critical detail. This novel approach to programming automation systems has been tested both in simulation and on prototype hardware.

Planar linear motors have been shown to be capable of fast accurate 2-DOF motions, making them useful for assembly tasks. However, the lack of sensing capability has limited their applications. Previous planar motor sensors were precise but bulky and sensed only a single direction of motion. We present a small integrated 3-DOF sensor capable of 0.2 /spl mu/m position resolution (1/spl sigma/) which has been integrated with an existing planar linear motor without changing its overall size. Physical design of the sensor is presented as well as the custom electronics used to produce a low-noise DC output from the small signal levels of the sensor. Finally, we present results that demonstrate the resolution and accuracy of the complete sensor system.

This paper presents a class of fundamental control policies suitable for use in a novel method for designing and specifying the dynamic motion of robotic systems. Through recourse to formal stability mechanisms a hybrid control strategy (one that relies on switching of underlying continuous policies) with desirable performance characteristics is demonstrated to be both stable and expressive for programming such motions. The long term intent is to utilize slightly more general control methods of this form to drastically simplify the process of integrating and programming modular automated assembly systems.

Abstract: "This paper serves to present our initial thoughts on the design of an architecture for highly flexible, modular, and distributed precision assembly systems. Through the use of a unified design, simulation, programming, and monitoring environment coupled with self- representing cooperative agents this architecture will significantly simplify the process of factory design and deployment. Our demonstration of this architecture takes the form of a 'table top' sized miniature assembly system targeted at the partial assembly of high-density magnetic storage devices."

We present a working implementation of a dynamics based architecture for visual sensing. This architecture provides field rate estimates of the positions and velocities of two independent falling balls in the face of repeated visual occlusions and departures from the field of view. The practical success of this system can be attributed to the interconnection of two strongly nonlinear dynamical systems: a novel triangulating state estimator; and an image plane window controller. We detail the architecture of this active sensor, provide data documenting its performance, and offer an analysis of its soundness in the form of a convergence proof for the estimator and a boundedness proof for the controller.

We report on our initial efforts to build robot feedback controllers that develop increased capability from simpler constituent controllers. Previous work with our three degree of freedom robot has resulted in a machine that exhibits various dynamically dexterous skills of superlative ability but very narrow behavioral scope. We focus here on the development of both a formalism and practice for the composition of constituent controllers. The composite should yield automatically purposive combinations of these skills that reach goals no one of the defining controllers could have achieved in isolation. The specific task we initially target, the "dynamical pick and place", requires the robot to acquire balls that have been "randomly" thrown into its work space and set them safely at rest in a specified location. We present a brief overview of the constituent behaviors and a mechanism for their combination along with documentation of our preliminary empirical successes.

This paper describes a simple visual servoing control algorithm capable of robustly positioning a three degree of freedom end effector based only on information from a stereo vision system. The proposed control algorithm does not require estimates of the gripper's spatial position, a significant source of calibration sensitivity. The controller is completely immune to positional camera calibration errors, and we demonstrate robustness to orientation miscalibration through a series of simulations and experiments.

In previous papers we have reported successful laboratory implementations of a family of juggling algorithms. In all but the one degree of freedom case, these empirically successful algorithms have so far resisted our analytical efforts to explain why they work. This is in large measure a consequence of our inability to write down using elementary functions an expression for the closed loop dynamics they induce. We discuss in this paper a modified juggling algorithm whose resulting closed loop dynamics can be written down directly. We offer data establishing the empirical success of the new algorithm. Theoretical analysis of the closed loop dynamics is presently in progress.>

In the recent successful effort to achieve the spatial two-juggle - batting two freely falling balls into independent stable periodic vertical orbits by repeated impacts with a three degree of freedom robot arm, the authors have found it necessary to introduce a dynamical window manager into their real-time stereo vision. This paper describes these necessary enhancements to the original vision system and then proposes a more formal account of how such a feedback based sensor might be understood to work. Further experimentation will be required to determine the extent to which the analytical model explains (and might thus be used as a tool to improve) the performance of the system presently working in the laboratory.

A model-based adaptive controller and proof of its global asymptotic stability with respect to the standard rigid-body model of robot-arm dynamics are presented. Experimental data from a study of one new and several established globally asymptotically stable adaptive controllers on two very different robot arms: (1) demonstrate the superior tracking performance afforded by the model-based algorithms over conventional PD control; (2) demonstrate and compare the superior performance of adaptive model-based algorithms over their nonadaptive counterparts; (3) reconcile several previous contrasting empirical studies; and (4) examine contexts that compromise their advantage.>

The authors review their progress to date in eliciting dynamically dexterous behaviors from a 3-d.o.f. direct drive robot manipulator whose real-time stereo cameras provide 60 Hz sampled images of multiple freely falling bodies in highly structured lighting conditions. At present, the robot is capable of forcing a single ping-pong ball into a specified steady-state (near) periodic vertical motion by repeated controlled impacts with a rigid paddle. The robot sustains the steady-state behavior over long periods (typically many thousands of impacts) and is capable of recovering from significant unexpected adversarial perturbations of the ball's flight phase. Gain tuning experiments corroborate the authors' contention that the stability mechanism underlying the robot's reliability can be attributed to the same nonlinear dynamics responsible for analogous behavior in a previous 1-d.o.f. robot.>

An adaptive controller is discussed and a proof of its global asymptotic stability with respect to the standard rigid body model of robot arm dynamics is presented. Experimental data from a study of this and other globally asymptotically stable adaptive controllers on two very different robot arms are used to reconcile several previous contrasting empirical studies, demonstrate and compare their superior tracking performance, and examine contexts that compromise their advantage.>