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

Results of ICRA 2011 Robot Challenge

  • A Modular Control System for Warehouse Automation - Algorithms and Simulations in USARSim Authors: Miklic, Damjan; Petrovic, Tamara; Coric, Mirko; Piskovic, Zvonimir; Bogdan, Stjepan
    In this paper, we present a control system for a fully autonomous material handling facility. The scenario we are considering is motivated by the 2011 IEEE Virtual Manufacturing Automation Challenge (VMAC). It consists of multiple autonomously guided vehicles (AGVs), transporting pallets of various goods between several input and output locations, through an unstructured warehouse environment. Only a map of the warehouse and a pallet delivery list are provided a priori. Pallets must be delivered to the output locations in the shortest time possible, while respecting the ordering of different pallet types specified by the delivery list. The presented control system handles all aspects of warehouse operation, from individual vehicle control to high-level mission planning and coordination. Delivery mission assignments are optimized using dynamic programming and simulated annealing techniques. Mission executions are coordinated using graph search methods and a modified version of the Banker's algorithm, to ensure safe, collision and deadlock-free system operation. System performance is evaluated on a virtual warehouse model, using the high fidelity USARSim simulator.
  • Wireless Swimming Microrobots: Design and Development of a 2 DoF Magnetic-Based System Authors: Palagi, Stefano; Lucarini, Gioia; Pensabene, Virginia; Levi, Alessandro; Mazzolai, Barbara; Menciassi, Arianna; Beccai, Lucia
    In this work, the design and development of an integrated platform for the steering of swimming microrobot is reported. The system consists of: a near-spherical soft and buoyant magnetic microrobot (with a diameter of about 500 µm) conceived for operation in liquid; a wireless magnetic steering system, including a compact magnetic field generator based on two pairs of Helmholtz and Maxwell coils; an electronic system for their driving; a control software; a joypad physical user interface; and, the micro-arena as working environment. The platform design fulfills the requirements for the “Mobility Task” of the 2011 NIST Mobile Microrobotics Challenge. The results obtained from preliminary validation experiments confirm that the microrobots can move in a fully controlled way, successfully accomplishing an intricate eight-shape path, as required, in the water filled micro-arena. In particular we achieved a maximum average speed of 0.71 mm/s and an exceptionally smooth motion.
  • Toward Fluidic Microrobots Using Electrowetting Authors: Schaler, Ethan; Tellers, Mary; Gerratt, Aaron P.; Penskiy, Ivan; Bergbreiter, Sarah
    This paper describes the performance of a fluidic microrobot using Electrowetting on Dielectric (EWOD). A system to control the fluidic microrobot was designed, constructed and deployed in the NIST Mobile Microrobotics Challenge at ICRA 2011. The microrobots (0.1 M KCl and 550 μm diameter) demonstrated the ability to perform controlled maneuvers in 2-D while transporting hydrophilic objects. The EWOD system is composed of a DIP-mounted die produced via standard microfabrication techniques and containing the control electrodes / competition arena, and a transparent ITO cover slip for grounding. Key advantages of this platform include a scalable design for batch EWOD system fabrication, potential simultaneous control of multiple microrobots, and an easily portable, compact system design.
  • A Textured Object Recognition Pipeline for Color and Depth Image Data Authors: Tang, Jie; Miller, Stephen; Singh, Arjun; Abbeel, Pieter
    We present an object recognition system which leverages the additional sensing and calibration information available in a robotics setting together with large amounts of training data to build high fidelity object models for a dataset of textured household objects. We then demonstrate how these models can be used for highly accurate detection and pose estimation in an end-to-end robotic perception system incorporating simultaneous segmentation, object classification, and pose fitting. The system can handle occlusions, illumination changes, multiple objects, and multiple instances of the same object. The system placed first in the ICRA 2011 Solutions in Perception instance recognition challenge. We believe the presented paradigm of building rich 3D models at training time and including depth information at test time is a promising direction for practical robotic perception systems.
  • The Jacobs Robotics Approach to Object Recognition and Localization in the Context of the ICRA'11 Solutions in Perception Challenge Authors: Vaskevicius, Narunas; Pathak, Kaustubh; Ichim, Alexandru-Eugen; Birk, Andreas
    In this paper, we give an overview of the Jacobs Robotics entry to the ICRA'11 Solutions in Perception Challenge. We present our multi-pronged strategy for object recognition and localization based on the integrated geometric and visual information available from the Kinect sensor. Firstly, the range image is over-segmented using an edge-detection algorithm and regions of interest are extracted based on a simple shape-analysis per segment. Then, these selected regions of the scene are matched with known objects using visual features and their distribution in 3D space. Finally, generated hypotheses about the positions of the objects are tested by back-projecting learned 3D models to the scene using estimated transformations and sensor model.

Teleoperation

  • Bilateral Teleoperation of Cooperative Manipulators Authors: Aldana, Carlos Iván; Nuno, Emmanuel; Basanez, Luis
    This paper presents an adaptive controller for the bilateral teleoperation of a system composed by a single local manipulator and multiple cooperative remote manipulators handling a common object. First, the nonlinear operational space dynamical behavior, of the complete teleoperation system, is derived and then, under the assumptions that the remote manipulators are rigidly grasping a non-deformable object and that the communications may induce constant time-delays, it is proved that velocities and position-orientation error between the local manipulator end-effector and the object asymptotically converge to zero. Simulations results are included to show the effectiveness of the proposed scheme.
  • Direct Force Reflecting Teleoperation with a Flexible Joint Robot Authors: Tobergte, Andreas; Albu-Schäffer, Alin
    This paper presents a high fidelity force feedback teleoperation control for surgical applications. Advanced control methods, such as flexible joint tracking control and passivity observation, are introduced in the direct force reflecting control architecture. A full state feedback controller of the flexible joint slave robot controls the motor position, velocity, the joint torque, and the torque derivative. The pose of the haptic device and the first three derivatives are observed to generate reference states for the robot control using the robot's inverse dynamics model. Interaction forces of the slave and the environment are measured with a force/torque sensor and directly sent back to the master device. Stability is guaranteed with a passivity observer that monitors the energy in the teleoperation system online and disconnects master and slave if the system operates beyond its stable region. The proposed control architecture is implemented with the sigma.7 haptic device and the MIRO robot. It is experimentally shown, that appropriately considering elasticities with full state reference and control of the slave, increases the dynamic range of the system enabling transparent and stable interaction with hard and soft environments.
  • Dynamic Scaling Interface for Assisted Teleoperation Authors: Munoz, Luis Miguel; Casals, Alicia
    Teleoperation, by adequately adapting computer interfaces, can benefit from the knowledge on human factors and psychomotor models in order to improve the effectiveness and efficiency in the execution of a task. While scaling is one of the performances frequently used in teleoperation tasks that require high precision, such as surgery, this article presents a scaling method that considers the system dynamics as well. The proposed dynamic scaling factor depends on the apparent position and velocity of the robot and targets. Such scaling improves the performance of teleoperation interfaces, thereby reducing user’s workload.
  • A Proportional Plus Damping Injection Controller for Teleoperators with Joint Flexibility and Time-Delays Authors: Nuno, Emmanuel; Sarras, Ioannis; Basanez, Luis; Kinnaert, Michel
    The problem of controlling a rigid bilateral teleoperator with time-delays has been effectively addressed since the late 80's. However, the control of flexible joint manipulators in a bilateral teleoperation scenario is still an open problem. In the present paper we report two versions of a proportional plus damping injection controller that are capable of globally stabilizing a nonlinear bilateral teleoperator with joint flexibility and variable time-delays. The first version controls a teleoperator composed by a rigid local manipulator and a flexible joint remote manipulator and the second version deals with local and remote manipulators with joint flexibility. For both schemes, it is proved that the joint and motor velocities and the local and remote position error are bounded. Moreover, if the human operator and remote environment forces are zero then velocities asymptotically converge to zero and position tracking is established. Simulations are presented to show the performance of the proposed controllers.
  • Stability of Position-Based Bilateral Telemanipulation Systems by Damping Injection Authors: Franken, Michel; Misra, Sarthak; Stramigioli, Stefano
    In this paper two different approaches to guaran- tee stability of bilateral telemanipulation systems are discussed. Both approaches inject damping into the system to guarantee passivity of the interaction with the device in the presence of time delays in the communication channel. The first approach derives tuning rules for a fixed viscous damper, whereas the second approach employs modulated dampers based upon the measured energy exchange with the device and enforces passivity in the time domain. Furthermore, a theoretical min- imum damping injection scheme is sketched that shows that the fixed damping approach is inherently conservative with respect to guaranteeing stability. Experimental results show that both the theoretical minimum damping scheme and a time domain passivity algorithm are successful in stabilizing the telemanipulation system for large time delays with lower gains of the damping elements than derived by the fixed damping injection approach. However, as damping is inherently present in the system, the tuning rules derived from the fixed damping injection approach can be used to identify if a time domain passivity algorithm is needed given boundary conditions on the actual time delays.
  • Bilateral Teleoperation of a Group of UAVs with Communication Delays and Switching Topology Authors: Secchi, Cristian; Franchi, Antonio; Buelthoff, Heinrich H.; Robuffo Giordano, Paolo
    In this paper, we present a passivity-based decentralized approach for bilaterally teleoperating a group of UAVs composing the slave side of the teleoperation system. In particular, we explicitly consider the presence of time delays, both among the master and slave, and within UAVs composing the group. Our focus is on analyzing suitable (passive) strategies that allow a stable teloperation of the group despite presence of delays, while still ensuring high flexibility to the group topology (e.g., possibility to autonomously split or join during the motion). The performance and soundness of the approach is validated by means of human/hardware-in-the-loop simulations (HHIL).

Applied Machine Learning

  • Active Learning from Demonstration for Robust Autonomous Navigation Authors: Silver, David; Bagnell, James; Stentz, Anthony
    Building robust and reliable autonomous navigation systems that generalize across environments and operating scenarios remains a core challenge in robotics. Machine learning has proven a significant aid in this task; in recent years learning from demonstration has become especially popular, leading to improved systems while requiring less expert tuning and interaction. However, these approaches still place a burden on the expert, specifically to choose the best demonstrations to provide. This work proposes two approaches for active learning from demonstration, in which the learning system requests specific demonstrations from the expert. The approaches identify examples for which expert demonstration is predicted to provide useful information on concepts which are either novel or uncertain to the current system. Experimental results demonstrate both improved generalization performance and reduced expert interaction when using these approaches.
  • Tendon-Driven Control of Biomechanical and Robotic Systems: A Path Integral Reinforcement Learning Approach Authors: Rombokas, Eric; Theodorou, Evangelos; Malhotra, Mark; Todorov, Emanuel; Matsuoka, Yoky
    We apply path integral reinforcement learning to a biomechanically accurate dynamics model of the index finger and then to the Anatomically Correct Testbed (ACT) robotic hand. We illustrate the applicability of Policy Improvement with Path Integrals to parameterized and non-parameterized control policies. This method is based on sampling variations in control, executing them in the real world, and minimizing a cost function on the resulting performance. Iteratively improving the control policy based on real-world performance requires no direct modeling of tendon network nonlinearities and contact transitions, allowing improved task performance.
  • Slip Prediction Using Hidden Markov Models: Multidimensional Sensor Data to Symbolic Temporal Pattern Learning Authors: Jamali, Nawid; Sammut, Claude
    We present experiments on the application of machine learning to predicting slip. The sensing information is provided by a force/torque sensor and an artificial finger, which has randomly distributed strain gauges and polyvinylidene fluoride (PVDF) films embedded in silicone resulting in multidimensional time series data on the finger-object contact. An incipient slip is detected by studying temporal patterns in the data. The data is analysed using probabilistic clustering that transforms the data into a sequence of symbols, which is used to train a hidden Markov model (HMM) classifier. Experimental results show that the classifier can predict a slip, at least 100ms before a slip takes place, with an accuracy of 96% on the validation set.
  • Collision-Free State Estimation Authors: Wong, Lawson L.S.; Kaelbling, Leslie; Lozano-Perez, Tomas
    In state estimation, we often want the maximum likelihood estimate of the current state. For the commonly used joint multivariate Gaussian distribution over the state space, this can be efficiently found using a Kalman filter. However, in complex environments the state space is often highly constrained. For example, for objects within a refrigerator, they cannot interpenetrate each other or the refrigerator walls. The multivariate Gaussian is unconstrained over the state space and cannot incorporate these constraints. In particular, the state estimate returned by the unconstrained distribution may itself be infeasible. Instead, we solve a related constrained optimization problem to find a good feasible state estimate. We illustrate this for estimating collision-free configurations for objects resting stably on a 2-D surface, and demonstrate its utility in a real robot perception domain.
  • Fault Detection and Isolation from Uninterpreted Data in Robotic Sensorimotor Cascades Authors: Censi, Andrea; Hakansson, Magnus; Murray, Richard
    One of the challenges in designing the next generation of robots operating in non-engineered environments is that there seems to be an infinite amount of causes that make the sensor data unreliable or actuators ineffective. In this paper, we discuss what faults are possible to detect using zero modeling effort: we start from uninterpreted streams of observations and commands, and without a prior knowledge of a model of the world. We show that in sensorimotor cascades it is possible to define static faults independently of a nominal model. We define an information-theoretic usefulness of a sensor reading and we show that it captures several kind of sensorimotor faults frequently encountered in practice. We particularize these ideas to the case of BDS/BGDS models, proposed in previous work as suitable candidates for describing generic sensorimotor cascades. We show several examples with camera and range-finder data, and we discuss a possible way to integrate these techniques in an existing robot software architecture.
  • Describing and Classifying Spatial and Temporal Contexts with OWL DL in Ubiquitous Robotics Authors: Sgorbissa, Antonio; Scalmato, Antonello; Zaccaria, Renato
    The article describes a system for describing and recognizing spatial and temporal patterns of events. The system is based on an ontology described through the Description Logics formalism and implemented in OWL DL. The approach is different from all other works in the literature since the system does not require an external reasoning engine, but relies only on the base mechanism for ontology classification. Experiments performed in two different scenarios are described, i.e., a Smart Home and a mobile robot for autonomous transportation operating within a partially automated building.