TechTalks from event: Technical session talks from ICRA 2012

Conference registration code to access these videos can be accessed by visiting this link: PaperPlaza. Step-by-step to access these videos are here: step-by-step process .
Why some of the videos are missing? If you had provided your consent form for your video to be published and still it is missing, please contact support@techtalks.tv

Grasping: Modeling, Analysis and Planning

  • On the Caging Region of a Third Finger with Object Boundary Clouds and Two Given Contact Positions Authors: Wan, Weiwei; Fukui, Rui; Shimosaka, Masamichi; Sato, Tomomasa
    This paper presents a caging approach which deals with planar boundary clouds collected from a laser scanner. Given the boundary clouds of a target object and two fixed finger positions, our aim is to find potential third finger positions that can prevent target from escaping into infinity. The major challenge in working with boundary clouds lies in their uncertainty in geometric model fitting and the failure of critical orientations. In this paper, we track canonical motions according to the rotational intersection of Configuration space fingers and rasterize Work space with grids to compute the third caging positions. Our approach can generate the capture region with max(O(np),O(h^2))<=O(n^2) cost where n denotes the resolution of grid rasterization, p denotes the resolution of canonical rasterization and h denotes the resolution of boundary rasterization or the number of boundary cloud points. Moreover, we propose a rough approximation which measures a subset of the possible positions by contracting rotations, indicating computational complexity of max(O(n),O(h^2)). In the experimental part, our proposal is compared with state-of-the-art works and applied to many other objects. The approach makes caging fast and effective.
  • Independent Contact Regions Based on a Patch Contact Model Authors: Charusta, Krzysztof Andrzej; Krug, Robert; Dimitrov, Dimitar Nikolaev; Iliev, Boyko
    The synthesis of multi-fingered grasps on non-trivial objects requires a realistic representation of the contact between the fingers of a robotic hand and an object. In this work, we use a patch contact model to approximate the contact between a rigid object and a deformable anthropomorphic finger. This contact model is utilized in the computation of Independent Contact Regions (ICRs) that have been proposed as a way to compensate for shortcomings in the finger positioning accuracy of robotic grasping devices. We extend the ICR algorithm to account for the patch contact model and show the benefits of this solution.
  • A Grasping Force Optimization Algorithm for Dexterous Robotic Hands Authors: Lippiello, Vincenzo; Siciliano, Bruno; Villani, Luigi
    The problem of grasping force optimization for a robotic system equipped with multi-fingered hands is considered in this paper. This problem is cast in a convex optimization problem, considering also joint torque constraints. A solution suitable for an online implementation, which allows a substantial reduction of the computational load by dynamically decreasing the number of active torque constraints is proposed. Moreover, for the case of a bimanual manipulation system, a sub-optimal single-hand optimization algorithm is presented and compared with the optimal one. The effectiveness of the described methods has been tested in a simulation case study.
  • Local Force Closure Authors: Kruger, Heinrich; Rimon, Elon; van der Stappen, Frank
    We introduce the concept of Local Force Closure. We define a local force closure grasp as a grasp which is capable of resisting some given external wrench as well as (through local variation in contact wrenches) any wrench in some neighborhood of the given wrench, with grasp quality exceeding some given threshold. Local force closure is useful in applications where a grasp only needs to resist some given external wrench, rather than fully constraining object, but where there is some uncertainty regarding the exact external wrench that needs to be resisted, or where there is a possibility of having to cope with some (relatively small) unknown disturbance forces. We show that by allowing disc-shaped fingers in contact with convex vertices of a polygonal object, any given wrench can be resisted by just two frictionless fingers. For a given polygonal object with <i>n</i> vertices and an external wrench <i>w</i><sub>ext</sub>, we show how to find all pairs of features of <i>P</i>, that admit grasps capable of resisting <i>w</i><sub>ext</sub> with grasp quality greater or equal to some threshold <i>Q</i>, in <i>O(n<sup>3/2+&#949;</sup>+K)</i> time, where <i>K</i> is the number of pairs in the output and <i>&#949;</i> is some arbitrarily small, positive constant. We then show how to adapt our algorithm to guarantee that the features reported, admit local force closure grasps.
  • Two-Fingered Caging of Polygons Via Contact-Space Graph Search Authors: Allen, Thomas F; Rimon, Elon; Burdick, Joel
    Based on a novel contact-space formulation, this paper presents a new algorithm to find two-fingered caging grasps of planar polygonal objects. We show that the caging problem has several useful properties in contact space. First, the critical points of the cage representation in the hand’s configuration space appear as critical points of an inter-finger distance function in contact space. Second, the critical points of this distance function can be simply characterized. Third, the contact space admits a rectangular decomposition where the distance function is convex in each rectangle, and all critical points lie on the rectangle boundaries. This property leads to a natural “caging graph,” which can be readily searched to construct the caging sets. An example, constructed from real-world data illustrates and validates the method.
  • Object Categorization and Grasping by Parts from Range Scan Data Authors: Aleotti, Jacopo; Lodi Rizzini, Dario; Caselli, Stefano
    Object category recognition and localization in 3D range data is of great importance in robot manipulation. In this work we propose a novel approach for object categorization and grasping that is focused on topological shape segmentation. The method allows generation of watertight triangulated models of the objects and their shape segmentation into parts. This segmentation provides meaningful information about grasp affordances. An efficient technique for encoding proximity data from range scans is also presented as well as an advanced strategy for manipulation of object sub-parts. Experiments are reported in a real environment using a robot arm equipped with eye-in-hand laser scanner and a parallel gripper.

Learning and Adaptive Control of Robotic Systems I

  • RTMBA: A Real-Time Model-Based Reinforcement Learning Architecture for Robot Control Authors: Hester, Todd; Quinlan, Michael; Stone, Peter
    Reinforcement Learning (RL) is a paradigm for learning decision-making tasks that could enable robots to learn and adapt to their situation on-line. For an RL algorithm to be practical for robotic control tasks, it must learn in very few samples, while continually taking actions in real-time. Existing model-based RL methods learn in relatively few samples, but typically take too much time between each action for practical on-line learning. In this paper, we present a novel parallel architecture for model-based RL that runs in real-time by 1) taking advantage of sample-based approximate planning methods and 2) parallelizing the acting, model learning, and planning processes in a novel way such that the acting process is sufficiently fast for typical robot control cycles. We demonstrate that algorithms using this architecture perform nearly as well as methods using the typical sequential architecture when both are given unlimited time, and greatly out-perform these methods on tasks that require real-time actions such as controlling an autonomous vehicle.
  • Sensorimotor Learning of Sound Localization from an Auditory Evoked Behavior Authors: Bernard, Mathieu; PIRIM, Patrick; de Cheveigné, Alain; Gas, Bruno
    A new method for self-supervised sensorimotor learning of sound source localization is presented, that allows a simulated listener to learn online an auditorimotor map from the sensorimotor experience provided by an auditory evoked behavior. The map represents the auditory space and is used to estimate the azimuthal direction of sound sources. The learning mainly consists in non-linear dimensionality reduction of sensorimotor data. Our results show that an auditorimotor map can be learned, both from real and simulated data, and that the online learning leads to accurate estimations of azimuthal sources direction.
  • Path-following Control of a Velocity Constrained Tracked Vehicle Incorporating Adaptive Slip Estimation Authors: Burke, Michael
    This work presents a model predictive path-following controller, which incorporates adaptive slip estimation for a tracked vehicle. Tracked vehicles are capable of manoeuvring in highly variable and uneven terrain, but difficulties in their control have traditionally limited their use as autonomous platforms. Attempts to compensate for slip in environments typically require that both the forward and rotational velocities of a platform be determined, but this can be challenging. This paper shows that it is possible to estimate vehicle traction using only a rate gyroscope, by providing a suitable adaptive least squares estimator to do so. An approach to generating slip compensating controls when platform velocity constraints are applied is also presented. The approach is controller independent, but we make use of a model predictive controller, vulnerable to the effects of model-plant mismatch, to highlight the efficacy of the proposed estimation and compensation. Path following results using a mixture model to generate feasible slip values are presented, and show a significant increase in controller performance.
  • Direct Yaw Moment Control for Four Wheel Independent Steering and Drive Vehicles Based on Centripetal Force Detection Authors: Lam, Tin Lun; Xu, Yangsheng
    In this paper, a deterministic yaw moment controller for four wheel independent steering and drive vehicles is proposed to enhance driving stability and controllability. Different to conventional methods that track a desired yaw rate, the proposed controller stabilizes a vehicle by additionally tracking the heading angle of a vehicle which is more efficient and robust. The heading angle of a vehicle is obtained by a novel method which is based on centripetal force detection. It eliminates the prerequisite knowledge of the characteristics between wheels and road surface which are time varying and difficult to be measured in real time. The proposed system only requires low cost sensing equipment such as wheel speed sensor and accelerometer that makes the system practical to be utilized. The proposed heading angle detection method can be generally applied to any kind of vehicle. The deterministic yaw moment controller is also applicable to any type of four wheel independent drive vehicles.
  • Predictive Control of Chained Systems: A Necessary Condition on the Control Horizon Authors: Courtial, Estelle; Fruchard, Matthieu; Allibert, Guillaume
    This paper deals with state feedback control of chained systems based on a Nonlinear Model Predictive Control (NMPC) strategy. Chained systems can model many common nonholonomic vehicles. We establish a relation between the degree of nonholonomy and the minimum length of the control horizon so as to make the NMPC feasible. A necessary condition on the control horizon of NMPC is given and theoretically proved whatever the dimension of the chained system consid- ered. This relation is used to design a NMPC-based control strategy for chained systems. One of the advantages of NMPC is the capability of taking into account the constraints on state and on control variables. The theoretical results are illustrated through simulations on a (2,5) chained system, describing a car-like vehicle with one trailer. Difficult motion objectives that require a lateral displacement are considered.
  • Xbots: An Approach to Generating and Executing Optimal Multi-Robot Plans with Cross-Schedule Dependencies Authors: Korsah, G. Ayorkor; Kannan, Balajee; Browning, Brett; Stentz, Anthony; Dias, M. Bernardine
    In this paper, we present an approach to bounded optimal planning and flexible execution for a robot team performing a set of spatially distributed tasks related by temporal ordering constraints such as precedence or synchronization. Furthermore, the manner in which the temporal constraints are satisfied impacts the overall utility of the team, due to the existence of both routing and delay costs. We present a bounded optimal offline planner for task allocation and scheduling in the presence of such cross-schedule dependencies, and a flexible, distributed online plan execution strategy. The integrated system performs task allocation and scheduling, executes the plans smoothly in the face of real-world variations in operation speed and task execution time, and ensures graceful degradation in the event of task failure. We demonstrate the capabilities of our approach on a team of three pioneer robots operating in an indoor environment. Experimental results demonstrate that approach is effective for constrained planning and execution in the face of real-world variations.

Marine Robotics I

  • Towards Improving Mission Execution for Autonomous Gliders with an Ocean Model and Kalman Filter Authors: Smith, Ryan N.; Kelly, Jonathan; Sukhatme, Gaurav
    Effective execution of a planned path by an underwater vehicle is important for proper analysis of the gathered science data, as well as to ensure the safety of the vehicle during the mission. Here, we propose the use of an unscented Kalman filter to aid in determining how the planned mission is executed. Given a set of waypoints that define a planned path and a dicretization of the ocean currents from a regional ocean model, we present an approach to determine the time interval at which the glider should surface to maintain a prescribed tracking error, while also limiting its time on the ocean surface. We assume practical mission parameters provided from previous field trials for the problem set up, and provide the simulated results of the Kalman filter mission planning approach. The results are initially compared to data from prior field experiments in which an autonomous glider executed the same path without pre-planning. Then, the results are validated through field trials with multiple autonomous gliders implementing different surfacing intervals simultaneously while following the same path.
  • Position and Velocity Filters for Intervention AUVs Based on Single Range and Depth Measurements Authors: Viegas, Daniel; Batista, Pedro; Oliveira, Paulo; Silvestre, Carlos
    This paper proposes novel cooperative navigation solutions for an Intervention Autonomous Underwater Vehicle (I-AUV) working in tandem with an Autonomous Surface Craft (ASC). The I-AUV is assumed to be moving in the presence of constant unknown ocean currents, and aims to estimate its position relying on measurements of its range to the ASC and of its depth relatively to the sea level. Two different scenarios are considered: in one, the ASC transmits its position and velocity to the I-AUV, while in the other the ASC transmits only its position, and the I-AUV has access to measurements of its velocity relative to the ASC. A sufficient condition for observability and a method for designing state observers with Globally Asymptotically Stable (GAS) error dynamics are presented for both problems. Finally, simulation results are included and discussed to assess the performance of the proposed solutions in the presence of measurement noise.
  • Uncertainty-Driven View Planning for Underwater Inspection Authors: Hollinger, Geoffrey; Englot, Brendan; Hover, Franz; Mitra, Urbashi; Sukhatme, Gaurav
    We discuss the problem of inspecting an underwater structure, such as a submerged ship hull, with an autonomous underwater vehicle (AUV). In such scenarios, the goal is to construct an accurate 3D model of the structure and to detect any anomalies (e.g., foreign objects or deformations). We propose a method for constructing 3D meshes from sonar-derived point clouds that provides watertight surfaces, and we introduce uncertainty modeling through non-parametric Bayesian regression. Uncertainty modeling provides novel cost functions for planning the path of the AUV to minimize a metric of inspection performance. We draw connections between the resulting cost functions and submodular optimization, which provides insight into the formal properties of active perception problems. In addition, we present experimental trials that utilize profiling sonar data from ship hull inspection.
  • Formation Control of Underactuated Autonomous Surface Vessels Using Redundant Manipulator Analogs Authors: Bishop, Bradley
    In this paper, we present a method utilizing redundant manipulator analogs for formation control of underactuated autonomous surface vessels (ASVs) with realistic turning constraints and dynamics. The method used relies on casting the swarm as a single entity and utilizing redundant manipulator techniques to guarantee task-level formation control as well as obstacle avoidance and secondary tasks such as mean position control. The method presented differs from other approaches in that the units herein represent a larger class of ASVs with realistic limitations on vessel motions and that the exact position of each of the units on the formation profile is not specified.
  • Delayed State Information Filter for USBL-Aided AUV Navigation Authors: Ribas, David; Ridao, Pere; Mallios, Angelos; Palomeras, Narcis
    This paper presents a navigation system for an Autonomous Underwater Vehicle (AUV) which merges standard dead reckoning navigation data with absolute position fixes from an Ultra-Short Base Line (USBL) system. Traditionally, the USBL transceiver is located on the surface, which makes necessary to feed the position fixes back to the AUV by means of an acoustic modem. An Information filter, which maintains a bounded circular buffer of past vehicle poses, is in charge of the sensor data fusion while dealing with de delays induced by the acoustic communication. The method is validated using a data set gathered for a dam inspection task.
  • Miniature Underwater Glider: Design, Modeling, and Experimental Results Authors: Zhang, Feitian; Thon, John; Thon, Cody; Tan, Xiaobo
    The concept of gliding robotic fish combines gliding and fin-actuation mechanisms to realize energy-efficient locomotion and high maneuverability, and holds strong promise for mobile sensing in versatile aquatic environments. In this paper we present the modeling and design of a miniature fish-like glider, a key enabling component for gliding robotic fish. The full dynamics of the glider is first derived and then reduced to the sagittal plane, where the lift, drag, and pitch moment coefficients are obtained as linear or quadratic functions of the attack angle based on computational fluid dynamics (CFD) analysis. The model is used to design the glider by accommodating stringent constraints on dimensions yet meeting the desired specification on speed. A fully untethered prototype of underwater glider is developed, with a weight of 4 kg and length of 40 cm. With a net buoyancy of 20 g, it realizes a steady gliding speed of 20 cm/s. The volume and net buoyancy of this glider are less than 10% and 5%, respectively, of those of reported gliders in the literature, and its speed per unit net buoyancy is over 9 times of those other vehicles. Experimental results have shown that the model is able to capture well both the steady glide behavior under different control inputs, and the dynamics during transients.