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

Stochastic Motion Planning

  • An Incremental Sampling-Based Algorithm for Stochastic Optimal Control Authors: Huynh, Vu Anh; Karaman, Sertac; Frazzoli, Emilio
    In this paper, we consider a class of continuous-time, continuous-space stochastic optimal control problems. Building upon recent advances in Markov chain approximation methods and sampling-based algorithms for deterministic path planning, we propose a novel algorithm called the incremental Markov Decision Process (iMDP) to compute incrementally control policies that approximate arbitrarily well an optimal policy in terms of the expected cost. The main idea behind the algorithm is to generate a sequence of finite discretizations of the original problem through random sampling of the state space. At each iteration, the discretized problem is a Markov Decision Process that serves as an incrementally refined model of the original problem. We show that with probability one, (i) the sequence of the optimal value functions for each of the discretized problems converges uniformly to the optimal value function of the original stochastic optimal control problem, and (ii) the original optimal value function can be computed efficiently in an incremental manner using asynchronous value iterations. Thus, the proposed algorithm provides an anytime approach to the computation of optimal control policies of the continuous problem. The effectiveness of the proposed approach is demonstrated on motion planning and control problems in cluttered environments in the presence of process noise.
  • Stochastic Distributed Multi-Agent Planning and Applications to Traffic Authors: Lim, Sejoon; Rus, Daniela
    This paper proposes a method for multi-agent path planning on a road network in the presence of congestion. We suggest a distributed method to find paths for multiple agents by introducing a probabilistic path choice achieving global goals such as the social optimum. This approach, which shows that the global goals can be achieved by local processing using only local information, can be parallelized and sped-up using massive parallel processing. The probabilistic assignment reliably copes with the case of random choices of unidentified agents or random route changes of agents who ignore our path guidance. We provide the analytical result on convergence and running time. We demonstrate and evaluate our algorithm by an implementation using asynchronous computation on multi-core computers.
  • Navigating between People: A Stochastic Optimization Approach Authors: Rios-Martinez, Jorge; Renzaglia, Alessandro; Spalanzani, Anne; Martinelli, Agostino; Laugier, Christian
    The objective of this paper is to present a strategy to safely move a robot in an unknown and complex environment where people are moving and interacting. The robot, by using only its sensor data, must navigate respecting humans’ comfort. To obtain good results in such a dynamic environment, a prediction on humans’ movement is also crucial. To solve all the aforementioned problems we introduce a suitable cost function. Its optimization is obtained by using a new stochastic and adaptive optimization algorithm (CAO). This method is very useful in particular when the analytical expression of the optimization function is unknown but numerical values are available for any state configuration. Additionally, the proposed method can easily incorporate any dynamical and environmental constraints. To validate the performance of the proposed solution, several simulation results are provided.
  • Probabilistic Path Planning for Multiple Robots with Subdimensional Expansion Authors: Wagner, Glenn; KANG, MINSU; Choset, Howie
    Probabilistic planners such as Rapidly-Exploring Random Trees (RRTs) and Probabilistic Roadmaps (PRMs) are powerful path planning algorithms for high dimensional systems, but even these potent techniques suffer from the curse of dimensionality, as can be seen in multirobot systems. In this paper, we apply a technique called subdimensional expansion in order to enhance the performance of probabilistic planners for multirobot path planning. We accomplish this by exploiting the structure inherent to such problems. Subdimensional expansion initially plans in each individual robot's configuration space separately. It then couples those spaces when robots come into close proximity with one another. In this way, we constrain a probabilistic planner to search a low dimensional space, while dynamically generating a higher dimensional space where necessary. We show in simulation that subdimensional expansion enhanced PRMs can solve problems involving 32 robots and 128 total degrees of freedom in less than 10 minutes. We also demonstrate that enhancing RRTs and PRMs with subdimensional expansion can decrease the time required to find a solution by more than an order of magnitude.
  • Stochastic Receding Horizon Control for Robots with Probabilistic State Constraints Authors: Shah, Shridhar; Pahlajani, Chetan; Lacock, Nicholaus; Tanner, Herbert G.
    This paper deals with the problem of receding horizon control of a robot subject to stochastic uncertainty within a constrained environment. We deviate from the conventional approach that minimizes expectation of a cost functional while ensuring satisfaction of probabilistic state constraints. Instead, we reduce the problem into a particular form of stochastic optimal control where the path that minimizes the cost functional is planned deterministically and a local stochastic optimal controller with exit constraints ensures satisfaction of probabilistic state constraints while following the planned path. This control design strategy ensures boundedness of errors around the reference path and collision-free convergence to the goal with probability one under the assumption of unbounded inputs. We show that explicit expressions for the control law are possible for certain cases. We provide simulation results for a point robot moving in a constrained two-dimensional environment under Brownian noise. The method can be extended to systems with bounded inputs, if a small nonzero probability of failure can be accepted.
  • High-Speed Flight in an Ergodic Forest Authors: Karaman, Sertac; Frazzoli, Emilio
    Inspired by birds flying through cluttered environments such as dense forests, this paper studies the theoretical foundations of high-speed motion through a randomly-generated obstacle field. Assuming that the locations and the sizes of the trees are determined by an ergodic point process, and under mild technical conditions on the dynamics of the bird, it is shown that the existence of an infinite collision-free trajectory through the forest exhibits a phase transition. In other words, if the bird flies faster than a certain critical speed, there is no infinite collision-free trajectory, with probability one, i.e., the bird will eventually collide with some tree, almost surely, regardless of the planning algorithm governing its motion. On the other hand, if the bird flies slower than this critical speed, then there exists at least one infinite collision-free trajectory, almost surely. Lower and upper bounds on the critical speed are derived for the special case of a Poisson forest considering a simple model for the bird's dynamics. Moreover, results from an extensive Monte-Carlo simulation study are presented. This paper also establishes novel connections between robot motion planning and statistical physics through ergodic theory and the theory of percolation, which may be of independent interest.

Medical Robotics II

  • Automatic Extraction of Command Hierarchies for Adaptive Brain-Robot Interfacing Authors: Bryan, Matthew; Nicoll, Griffin; Thomas, Vibinash; Chung, Mike; Smith, Joshua R.; Rao, Rajesh P. N.
    Recent advances in neuroscience and robotics have allowed initial demonstrations of brain-computer interfaces (BCIs) for controlling wheeled and humanoid robots. However, further advances have proved challenging due to the low throughput of the interfaces and the high degrees-of-freedom (DOF) of the robots. In this paper, we build on our previous work on Hierarchical BCIs (HBCIs) which seek to mitigate this problem. We extend HBCIs to allow training of arbitrarily complex tasks, with training no longer restricted to a particular robot state space (such as Cartesian space for a navigation task). We present two algorithms for learning command hierarchies by automatically extracting patterns from a user's command history. The first algorithm builds an arbitrary-level hierarchical structure (a "control grammar") whose elements can represent skills, whole tasks, collections of tasks, etc. The user "executes" single symbols from this grammar, which produce sequences of lower-level commands. The second algorithm, which is probabilistic, also learns sequences which can be executed as high-level commands, but does not build an explicit hierarchical structure. Both algorithms provide a de facto form of dictionary compression, which enhances the effective throughput of the BCI. We present results from two human subjects who successfully used the hierarchical BCI to control a simulated PR2 robot using brain signals recorded non-invasively through electroencephalography (EEG).
  • Powered Wheelchair Navigation Assistance through Kinematically Correct Environmental Haptic Feedback Authors: Vander Poorten, Emmanuel B; Demeester, Eric; Reekmans, Eli; Huntemann, Alexander; De Schutter, Joris
    This article introduces a set of novel haptic guidance algorithms intended to provide intuitive and reliable assistance for electric wheelchair navigation through narrow or crowded spaces. The proposed schemes take hereto the nonholonomic nature and a detailed geometry of the wheelchair into consideration. The methods encode the environment as a set of collision-free circular paths and, making use of a model-free impedance controller, ‘haptically’ guide the user along collision-free paths or away from obstructed paths or paths that simply do not coincide with the motion intended by the user. The haptic feedback plays a central role as it establishes a fast bilateral communication channel between user and wheelchair controller and allows a direct negotiation about wheelchair motion. If found unsatisfactory, suggested trajectories can always be overruled by the user. Relying on inputs from user modeling and intention recognition schemes, the system can reduce forces needed to move along intended directions, thereby avoiding unnecessary fatigue of the user. A commercial powered wheelchair was upgraded and feasability tests were conducted to validate the proposed methods. The potential of the proposed approaches was hereby demonstrated.
  • A Haptic Instruction Based Assisted Driving System for Training the Reverse Parking Authors: Hirokawa, Masakazu; Uesugi, Naohisa; Furugori, Satoru; Kitagawa, Tomoko; Suzuki, Kenji
    The accident probability of beginner drivers is significantly higher than that of experienced drivers. It can be assumed that this is due to lack of driving skills which lead to making wrong decisions according to cognition and operating in correct way. In this paper, we propose a novel assisted driving system intended to help drivers to improve their skills for the reverse parking. The system is able to assist the driver by haptic instruction on the steering wheel in order to induce the driver to make the adequate operation. For the validation, we developed a 1/10 scale car simulator as a simulation environment on which we installed the proposed assistance method and conducted reverse parking experiment by using the simulator. According to the experiment, we validated that the parking accuracy and the trajectory similarity of subjects assisted by proposed system significantly increased compare to subjects unassisted. Consequently, the proposed assisted driving system could accelerate the learning of humans' driving skills.
  • Remote Palpation to Localize Tumors in Robot-Assisted Minimally Invasive Approach Authors: Talasaz, Ali; Patel, Rajnikant V.
    This paper presents a new tactile-force integrated method to localize tumors minimally invasively using robotic assistance. This method relies on using a capacitive sensor at the tip of a Tactile Sensing Instrument (TSI) which can be inserted into a patient's body in a minimally invasive manner. In this work, the operator palpates tissue containing tumors in a minimally invasive surgical (MIS) training box, representing the patient's body, through a master-slave teleoperation system which consists of a 7 degrees-of-freedom (DOF) haptic interface, used as the master, and a Mitsubishi PA10-7C robot as the slave. Using the proposed method, the operator would be able to palpate the tissue consistently, observe the pressure distribution over the tissue by a color contour map on a screen and feel the tumor on his/her fingers through a grasping mechanism of the haptic interface as a result of higher stiffness of the tumor. The tissue used for the experiments was ex vivo bovine lung and seven participants were asked to locate artificial tumors embedded in the lungs. The results show an accuracy of 93% in tumor localization using the proposed method while the average force applied to the tissue was 3.42 N and the force never exceeded 6 N.
  • Improvements in the Control of a Flexible Endoscopic System Authors: Bardou, Berengere; Nageotte, Florent; zanne, Philippe; de Mathelin, Michel
    The use of flexible cable-driven systems is common in medicine (endoscope, catheter...). Their flexiblity allows surgeons to reach internal organs through sinuous and constrained ways. Unfortunately these systems are subject to backlash due to their internal mechanism. These non linearities raise many difficulties when robotizing and controlling such systems. In this article we propose an approach to improve the cartesian control of a four ways flexible endoscopic system with strong and unknown backlash-like non linearities. The method is based on an automatic off-line hystereses learning. We show that, despite coupling between degrees of freedom, it is possible to extract information from the hystereses which allow to improve cartesian control. Experiments on a real endoscopic system show the validity and the interest of the approach.

Vision-Based Attention and Interaction

  • Computing Object-Based Saliency in Urban Scenes Using Laser Sensing Authors: Zhao, Yipu; He, Mengwen; Zhao, Huijing; Davoine, Franck; Zha, Hongbin
    It becomes a well-known technology that a low-level map of complex environment containing 3D laser points can be generated using a robot with laser scanners. Given a cloud of 3D laser points of an urban scene, this paper proposes a method for locating the objects of interest, e.g. traffic signs or road lamps, by computing object-based saliency. Our major contributions are: 1) a method for extracting simple geometric features from laser data is developed, where both range images and 3D laser points are analyzed; 2) an object is modeled as a graph used to describe the composition of geometric features; 3) a graph matching based method is developed to locate the objects of interest on laser data. Experimental results on real laser data depicting urban scenes are presented; efficiency as well as limitations of the method are discussed.
  • Where Do I Look Now? Gaze Allocation During Visually Guided Manipulation Authors: Nunez-Varela, Jose; Ravindran, Balaraman; Wyatt, Jeremy
    In this work we present principled methods for the coordination of a robot's oculomotor system with the rest of its body motor systems. The problem is to decide which physical actions to perform next and where the robot's gaze should be directed in order to gain information that is relevant to the success of its physical actions. Previous work on this problem has shown that a reward-based coordination mechanism provides an efficient solution. However, that approach does not allow the robot to move its gaze to different parts of the scene, it considers the robot to have only one motor system, and assumes that the actions have the same duration. The main contributions of our work are to extend that previous reward-based approach by making decisions about where to fixate the robot's gaze, handling multiple motor systems, and handling actions of variable duration. We compare our approach against two common baselines: random and round robin gaze allocation. We show how our method provides a more effective strategy to allocate gaze where it is needed the most.
  • 3D AAM Based Face Alignment under Wide Angular Variations Using 2D and 3D Data Authors: Wang, Chieh-Chih
    Active Appearance Models (AAMs) are widely used to estimate the shape of the face together with its orientation, but AAM approaches tend to fail when the face is under wide angular variations. Although it is feasible to capture the overall 3D face structure using 3D data from range cameras, the locations of facial features are often estimated imprecisely or incorrectly due to depth measurement uncertainty. Face alignment using 2D and 3D images suffer from different issues and have varying reliability in different situations. The existing approaches introduce a weighting function to balance 2D and 3D alignments in which the weighting function is tuned manually and the sensor characteristics are not taken into account. In this paper, we propose to balance 3D face alignment using 2D and 3D data based on the observed data and the sensors characteristics. The feasibility of wide-angle face alignment is demonstrated using two different sets of depth and conventional cameras. The experimental results show that a stable alignment is achieved with a maximum improvement of 26% compared to 3D AAM using 2D image and 30% improvement over the state-of-the-art 3DMM methods in terms of 3D head pose estimation.
  • Robots That Validate Learned Perceptual Models Authors: Klank, Ulrich; Mösenlechner, Lorenz; Maldonado, Alexis; Beetz, Michael
    Service robots that should operate autonomously need to perform actions reliably, and be able to adapt to their changing environment using learning mechanisms. Optimally, robots should learn continuously but this approach often suffers from problems like over-fitting, drifting or dealing with incomplete data. In this paper, we propose a method to automatically validate autonomously acquired perception models. These perception models are used to localize objects in the environment with the intention of manipulating them with the robot. Our approach verifies the learned perception models by moving the robot, trying to re-detect an object and then to grasp it. From observable failures of these actions and high-level loop-closures to validate the eventual success, we can derive certain qualities of our models and our environment. We evaluate our approach by using two different detection algorithms, one using 2D RGB data and one using 3D point clouds. We show that our system is able to improve the perception performance significantly by learning which of the models is better in a certain situation and a specific context. We show how additional validation allows for successful continuous learning. The strictest precondition for learning such perceptual models is correct segmentation of objects which is evaluated in a second experiment.
  • Uncalibrated Visual Servoing for Intuitive Human Guidance of Robots Authors: Marshall, Matthew; Matthews, James; Hu, Ai-Ping; McMurray, Gary
    We propose a novel implementation of visual servoing whereby a human operator can guide a robot relative to the coordinate frame of an eye-in-hand camera. Among other applications, this can allow the operator to work in the image space of the eye-in-hand camera. This is achieved using a gamepad, a time-of-flight camera (an active sensor that creates depth data), and recursive least-squares update with Gauss-Newton control. Contributions of this paper include the use of a person to cause the control action in a visual-servoing system, and the introduction of uncalibrated position-based visual servoing. The system's efficacy is evaluated via trials involving human operators in different scenarios.
  • Leveraging RGB-D Data: Adaptive Fusion and Domain Adaptation for Object Detection Authors: Spinello, Luciano; Luber, Matthias; Arras, Kai Oliver
    Vision and range sensing belong to the richest sensory modalities for perception in robotics and related fields. This paper addresses the problem of how to best combine image and range data for the task of object detection. In particular, we propose a novel adaptive fusion approach, hierarchical Gaussian Process mixtures of experts, able to account for missing information and cross-cue data consistency. The hierarchy is a two-tier architecture that for each modality, each frame and each detection computes a weight function using Gaussian Processes that reflects the confidence of the respective information. We further propose a method called cross-cue domain adaptation that makes use of large image data sets to improve the depth-based object detector for which only few training samples exist. In the experiments that include a comparison with alternative sensor fusion schemes, we demonstrate the viability of the proposed methods and achieve significant improvements in classification accuracy.