Formal methods for control synthesis in partially observed environments : application to autonomous robotic manipulation

Modern robots are increasingly expected to function in uncertain and dynamically challenging environments, often in proximity with humans. In addition, wide scale adoption of robots requires on-the-fly adaptability of software for diverse application. These requirements strongly suggest the need to adopt formal representations of high level goals and safety specifications, especially as temporal logic formulas. This approach allows for the use of formal verification techniques for controller synthesis that can give guarantees for safety and performance. Robots operating in unstructured environments also face limited sensing capability. Correctly inferring a robot's progress toward high level goal can be challenging.

This thesis develops new algorithms for synthesizing discrete controllers in partially known environments under specifications represented as linear temporal logic (LTL) formulas. It is inspired by recent developments in finite abstraction techniques for hybrid systems and motion planning problems. The robot and its environment is assumed to have a finite abstraction as a Partially Observable Markov Decision Process (POMDP), which is a powerful model class capable of representing a wide variety of problems. However, synthesizing controllers that satisfy LTL goals over POMDPs is a challenging problem which has received only limited attention.

This thesis proposes tractable, approximate algorithms for the control synthesis problem using Finite State Controllers (FSCs). The use of FSCs to control finite POMDPs allows for the closed system to be analyzed as finite global Markov chain. The thesis explicitly shows how transient and steady state behavior of the global Markov chains can be related to two different criteria with respect to satisfaction of LTL formulas. First, the maximization of the probability of LTL satisfaction is related to an optimization problem over a parametrization of the FSC. Analytic computation of gradients are derived which allows the use of first order optimization techniques.

The second criterion encourages rapid and frequent visits to a restricted set of states over infinite executions. It is formulated as a constrained optimization problem with a discounted long term reward objective by the novel utilization of a fundamental equation for Markov chains - the Poisson equation. A new constrained policy iteration technique is proposed to solve the resulting dynamic program, which also provides a way to escape local maxima.

The algorithms proposed in the thesis are applied to the task planning and execution challenges faced during the DARPA Autonomous Robotic Manipulation - Software challenge.

Optimization-Based Robotic Manipulation for Safe Interaction with Human Operators

This thesis investigates a method for human-robot interaction (HRI) in order to uphold productivity of industrial robots like minimization of the shortest operation time, while ensuring human safety like collision avoidance. For solving such problems an online motion planning approach for robotic manipulators with HRI has been proposed. The approach is based on model predictive control (MPC) with embedded mixed integer programming. The planning strategies of the robotic manipulators mainly considered in the thesis are directly performed in the workspace for easy obstacle representation. The non-convex optimization problem is approximated by a mixed-integer program (MIP). It is further effectively reformulated such that the number of binary variables and the number of feasible integer solutions are drastically decreased. Safety-relevant regions, which are potentially occupied by the human operators, can be generated online by a proposed method based on hidden Markov models. In contrast to previous approaches, which derive predictions based on probability density functions in the form of single points, such as most likely or expected human positions, the proposed method computes safety-relevant subsets of the workspace as a region which is possibly occupied by the human at future instances of time. The method is further enhanced by combining reachability analysis to increase the prediction accuracy. These safety-relevant regions can subsequently serve as safety constraints when the motion is planned by optimization. This way one arrives at motion plans that are safe, i.e. plans that avoid collision with a probability not less than a predefined threshold. The developed methods have been successfully applied to a developed demonstrator, where an industrial robot works in the same space as a human operator. The task of the industrial robot is to drive its end-effector according to a nominal sequence of grippingmotion-releasing operations while no collision with a human arm occurs.

Visual perception for the 3D recognition of geometric pieces in robotic manipulation

During grasping and intelligent robotic manipulation tasks, the camera position relative to the scene changes dramatically because the robot is moving to adapt its path and correctly grasp objects. This is because the camera is mounted at the robot effector. For this reason, in this type of environment, a visual recognition system must be implemented to recognize and “automatically and autonomously” obtain the positions of objects in the scene. Furthermore, in industrial environments, all objects that are manipulated by robots are made of the same material and cannot be differentiated by features such as texture or color. In this work, first, a study and analysis of 3D recognition descriptors has been completed for application in these environments. Second, a visual recognition system designed from specific distributed client-server architecture has been proposed to be applied in the recognition process of industrial objects without these appearance features. Our system has been implemented to overcome problems of recognition when the objects can only be recognized by geometric shape and the simplicity of shapes could create ambiguity. Finally, some real tests are performed and illustrated to verify the satisfactory performance of the proposed system.

Towards vision-based deep reinforcement learning for robotic motion control

This paper introduces a machine learning based system for controlling a robotic manipulator with visual perception only. The capability to autonomously learn robot controllers solely from raw-pixel images and without any prior knowledge of configuration is shown for the first time. We build upon the success of recent deep reinforcement learning and develop a system for learning target reaching with a three-joint robot manipulator using external visual observation. A Deep Q Network (DQN) was demonstrated to perform target reaching after training in simulation. Transferring the network to real hardware and real observation in a naive approach failed, but experiments show that the network works when replacing camera images with synthetic images.

Estimation and inference for grasping and manipulation tasks using vision and kinesthetic sensors

This thesis presents a novel framework for state estimation in the context of robotic grasping and manipulation. The overall estimation approach is based on fusing various visual cues for manipulator tracking, namely appearance and feature-based, shape-based, and silhouette-based visual cues. Similarly, a framework is developed to fuse the above visual cues, but also kinesthetic cues such as force-torque and tactile measurements, for in-hand object pose estimation. The cues are extracted from multiple sensor modalities and are fused in a variety of Kalman filters.

A hybrid estimator is developed to estimate both a continuous state (robot and object states) and discrete states, called contact modes, which specify how each finger contacts a particular object surface. A static multiple model estimator is used to compute and maintain this mode probability. The thesis also develops an estimation framework for estimating model parameters associated with object grasping. Dual and joint state-parameter estimation is explored for parameter estimation of a grasped object's mass and center of mass. Experimental results demonstrate simultaneous object localization and center of mass estimation.

Dual-arm estimation is developed for two arm robotic manipulation tasks. Two types of filters are explored; the first is an augmented filter that contains both arms in the state vector while the second runs two filters in parallel, one for each arm. These two frameworks and their performance is compared in a dual-arm task of removing a wheel from a hub.

This thesis also presents a new method for action selection involving touch. This next best touch method selects an available action for interacting with an object that will gain the most information. The algorithm employs information theory to compute an information gain metric that is based on a probabilistic belief suitable for the task. An estimation framework is used to maintain this belief over time. Kinesthetic measurements such as contact and tactile measurements are used to update the state belief after every interactive action. Simulation and experimental results are demonstrated using next best touch for object localization, specifically a door handle on a door. The next best touch theory is extended for model parameter determination. Since many objects within a particular object category share the same rough shape, principle component analysis may be used to parametrize the object mesh models. These parameters can be estimated using the action selection technique that selects the touching action which best both localizes and estimates these parameters. Simulation results are then presented involving localizing and determining a parameter of a screwdriver.

Lastly, the next best touch theory is further extended to model classes. Instead of estimating parameters, object class determination is incorporated into the information gain metric calculation. The best touching action is selected in order to best discern between the possible model classes. Simulation results are presented to validate the theory.

Control framework for dexterous manipulation using dynamic visual servoing and tactile sensors’ feedback

Tactile sensors play an important role in robotics manipulation to perform dexterous and complex tasks. This paper presents a novel control framework to perform dexterous manipulation with multi-fingered robotic hands using feedback data from tactile and visual sensors. This control framework permits the definition of new visual controllers which allow the path tracking of the object motion taking into account both the dynamics model of the robot hand and the grasping force of the fingertips under a hybrid control scheme. In addition, the proposed general method employs optimal control to obtain the desired behaviour in the joint space of the fingers based on an indicated cost function which determines how the control effort is distributed over the joints of the robotic hand. Finally, authors show experimental verifications on a real robotic manipulation system for some of the controllers derived from the control framework.

Human-robot interaction for telemanipulation in large workspaces

Independientemente de la existencia de técnicas altamente sofisticadas y capacidades de cómputo cada vez más elevadas, los problemas asociados a los robots que interactúan con entornos no estructurados siguen siendo un desafío abierto en robótica. A pesar de los grandes avances de los sistemas robóticos autónomos, hay algunas situaciones en las que una persona en el bucle sigue siendo necesaria. Ejemplos de esto son, tareas en entornos de fusión nuclear, misiones espaciales, operaciones submarinas y cirugía robótica. Esta necesidad se debe a que las tecnologías actuales no pueden realizar de forma fiable y autónoma cualquier tipo de tarea. Esta tesis presenta métodos para la teleoperación de robots abarcando distintos niveles de abstracción que van desde el control supervisado, en el que un operador da instrucciones de alto nivel en la forma de acciones, hasta el control bilateral, donde los comandos toman la forma de señales de control de bajo nivel. En primer lugar, se presenta un enfoque para llevar a cabo la teleoperación supervisada de robots humanoides. El objetivo es controlar robots terrestres capaces de ejecutar tareas complejas en entornos de búsqueda y rescate utilizando enlaces de comunicación limitados. Esta propuesta incorpora comportamientos autónomos que el operador puede utilizar para realizar tareas de navegación y manipulación mientras se permite cubrir grandes áreas de entornos remotos diseñados para el acceso de personas. Los resultados experimentales demuestran la eficacia de los métodos propuestos. En segundo lugar, se investiga el uso de dispositivos rentables para telemanipulación guiada. Se presenta una aplicación que involucra un robot humanoide bimanual y un traje de captura de movimiento basado en sensores inerciales. En esta aplicación, se estudian las capacidades de adaptación introducidas por el factor humano y cómo estas pueden compensar la falta de sistemas robóticos de alta precisión. Este trabajo es el resultado de una colaboración entre investigadores del Biorobotics Laboratory de la Universidad de Harvard y el Centro de Automática y Robótica UPM-CSIC. En tercer lugar, se presenta un nuevo controlador háptico que combina velocidad y posición. Este controlador bilateral híbrido hace frente a los problemas relacionados con la teleoperación de un robot esclavo con un gran espacio de trabajo usando un dispositivo háptico pequeño como maestro. Se pueden cubrir amplias áreas de trabajo al cambiar automáticamente entre los modos de control de velocidad y posición. Este controlador háptico es ideal para sistemas maestro-esclavo con cinemáticas diferentes, donde los comandos se transmiten en el espacio de la tarea del entorno remoto. El método es validado para realizar telemanipulación hábil de objetos con un robot industrial. Por último, se introducen dos contribuciones en el campo de la manipulación robótica. Por un lado, se presenta un nuevo algoritmo de cinemática inversa, llamado método iterativo de desacoplamiento cinemático. Este método se ha desarrollado para resolver el problema cinemático inverso de un tipo de robot de seis grados de libertad donde una solución cerrada no está disponible. La eficacia del método se compara con métodos numéricos convencionales. Además, se ha diseñado una taxonomía robusta de agarres que permite controlar diferentes manos robóticas utilizando una correspondencia, basada en gestos, entre los espacios de trabajo de la mano humana y de la mano robótica. El gesto de la mano humana se identifica mediante la lectura de los movimientos relativos del índice, el pulgar y el dedo medio del usuario durante las primeras etapas del agarre. ABSTRACT Regardless of the availability of highly sophisticated techniques and ever increasing computing capabilities, the problems associated with robots interacting with unstructured environments remains an open challenge. Despite great advances in autonomous robotics, there are some situations where a humanin- the-loop is still required, such as, nuclear, space, subsea and robotic surgery operations. This is because the current technologies cannot reliably perform all kinds of task autonomously. This thesis presents methods for robot teleoperation strategies at different levels of abstraction ranging from supervisory control, where the operator gives high-level task actions, to bilateral teleoperation, where the commands take the form of low-level control inputs. These strategies contribute to improve the current human-robot interfaces specially in the case of slave robots deployed at large workspaces. First, an approach to perform supervisory teleoperation of humanoid robots is presented. The goal is to control ground robots capable of executing complex tasks in disaster relief environments under constrained communication links. This proposal incorporates autonomous behaviors that the operator can use to perform navigation and manipulation tasks which allow covering large human engineered areas of the remote environment. The experimental results demonstrate the efficiency of the proposed methods. Second, the use of cost-effective devices for guided telemanipulation is investigated. A case study involving a bimanual humanoid robot and an Inertial Measurement Unit (IMU) Motion Capture (MoCap) suit is introduced. Herein, it is corroborated how the adaptation capabilities offered by the human-in-the-loop factor can compensate for the lack of high-precision robotic systems. This work is the result of collaboration between researchers from the Harvard Biorobotics Laboratory and the Centre for Automation and Robotics UPM-CSIC. Thirdly, a new haptic rate-position controller is presented. This hybrid bilateral controller copes with the problems related to the teleoperation of a slave robot with large workspace using a small haptic device as master. Large workspaces can be covered by automatically switching between rate and position control modes. This haptic controller is ideal to couple kinematic dissimilar master-slave systems where the commands are transmitted in the task space of the remote environment. The method is validated to perform dexterous telemanipulation of objects with a robotic manipulator. Finally, two contributions for robotic manipulation are introduced. First, a new algorithm, the Iterative Kinematic Decoupling method, is presented. It is a numeric method developed to solve the Inverse Kinematics (IK) problem of a type of six-DoF robotic arms where a close-form solution is not available. The effectiveness of this IK method is compared against conventional numerical methods. Second, a robust grasp mapping has been conceived. It allows to control a wide range of different robotic hands using a gesture based correspondence between the human hand space and the robotic hand space. The human hand gesture is identified by reading the relative movements of the index, thumb and middle fingers of the user during the early stages of grasping.

Object slippage prevention using different control strategies

This paper presents experimental results obtained using a friction-based slippage and tangential force sensing device that has been developed for the purpose of reliable object slippage prevention during robotic manipulation. The experimental results obtained demonstrate that the developed slippage sensing strategy is rugged and reliable even in its current “rough prototype” state of development. This work has the potential to yield a low cost and highly customisable slippage and tangential force sensing device for a variety of robotic object grasping and manipulation applications. It is envisaged that the work presented here will be beneficial to researchers in the area of object slippage prevention.

Robotic object grasping in context of human grasping and manipulation

This paper presents experimental and deductive findings that shed new light on grasp force estimation, which improves robot’s chances to grasp and manipulate the object close to optimum conditions on the first attempt, which in turn improves robot’s object manipulation dexterity.
This paper proposes that object slippage detection in the human hand is not detected based purely on microvibrations sensed by the human skin during incipient slippage but also on load sensing at each finger and movement of fingers relative to each other while holding an object.

Robotic grasping and manipulation controller framework: architecture redevelopment

This paper details the further improvements obtained by redesigning a previously offered Manipulation Controller Framework to provide support to an innovative, friction-based object slippage detection strategy employed by the robotic object manipulator. This upgraded Manipulation Controller Framework includes improved slippage detection functionality and a streamlined architecture designed to improve controller robustness, reliability and speed. Improvements include enhancements to object slippage detection strategy, the removal of the decision making module and integration of its functionality into the Motion Planner, and the stream-lining of the Motion Planner to improve its effectiveness. It is anticipated that this work will be useful to researchers developing integrated robot controller architectures and slippage control.

Integrated visual perception architecture for robotic clothes perception and manipulation

This thesis proposes a generic visual perception architecture for robotic clothes perception and manipulation. This proposed architecture is fully integrated with a stereo vision system and a dual-arm robot and is able to perform a number of autonomous laundering tasks. Clothes perception and manipulation is a novel research topic in robotics and has experienced rapid development in recent years. Compared to the task of perceiving and manipulating rigid objects, clothes perception and manipulation poses a greater challenge. This can be attributed to two reasons: firstly, deformable clothing requires precise (high-acuity) visual perception and dexterous manipulation; secondly, as clothing approximates a non-rigid 2-manifold in 3-space, that can adopt a quasi-infinite configuration space, the potential variability in the appearance of clothing items makes them difficult to understand, identify uniquely, and interact with by machine. From an applications perspective, and as part of EU CloPeMa project, the integrated visual perception architecture refines a pre-existing clothing manipulation pipeline by completing pre-wash clothes (category) sorting (using single-shot or interactive perception for garment categorisation and manipulation) and post-wash dual-arm flattening. To the best of the author’s knowledge, as investigated in this thesis, the autonomous clothing perception and manipulation solutions presented here were first proposed and reported by the author. All of the reported robot demonstrations in this work follow a perception-manipulation method- ology where visual and tactile feedback (in the form of surface wrinkledness captured by the high accuracy depth sensor i.e. CloPeMa stereo head or the predictive confidence modelled by Gaussian Processing) serve as the halting criteria in the flattening and sorting tasks, respectively. From scientific perspective, the proposed visual perception architecture addresses the above challenges by parsing and grouping 3D clothing configurations hierarchically from low-level curvatures, through mid-level surface shape representations (providing topological descriptions and 3D texture representations), to high-level semantic structures and statistical descriptions. A range of visual features such as Shape Index, Surface Topologies Analysis and Local Binary Patterns have been adapted within this work to parse clothing surfaces and textures and several novel features have been devised, including B-Spline Patches with Locality-Constrained Linear coding, and Topology Spatial Distance to describe and quantify generic landmarks (wrinkles and folds). The essence of this proposed architecture comprises 3D generic surface parsing and interpretation, which is critical to underpinning a number of laundering tasks and has the potential to be extended to other rigid and non-rigid object perception and manipulation tasks. The experimental results presented in this thesis demonstrate that: firstly, the proposed grasp- ing approach achieves on-average 84.7% accuracy; secondly, the proposed flattening approach is able to flatten towels, t-shirts and pants (shorts) within 9 iterations on-average; thirdly, the proposed clothes recognition pipeline can recognise clothes categories from highly wrinkled configurations and advances the state-of-the-art by 36% in terms of classification accuracy, achieving an 83.2% true-positive classification rate when discriminating between five categories of clothes; finally the Gaussian Process based interactive perception approach exhibits a substantial improvement over single-shot perception. Accordingly, this thesis has advanced the state-of-the-art of robot clothes perception and manipulation.