Skip to main content
main-content

Über dieses Buch

This book constitutes the refereed proceedings of the Third International Conference on Simulation, Modeling, and Programming for Autonomous Robots, SIMPAR 2012, held in Tsukuba, Japan, in November 2012. The 33 revised full papers and presented together with 3 invited talks were carefully reviewed and selected from 46 submissions. Ten papers describe design of complex behaviors of autonomous robots, 9 address software layers, 8 papers refer to related modeling and learning. The papers are organized in topical sections on mobile robots, software modeling and architecture and humanoid and biped robots.

Inhaltsverzeichnis

Frontmatter

Invited Talks

A Geometric Perspective of Anthropomorphic Embodied Actions

Starting from a mechanics point of view, the human (or humanoid) body is both a redundant system and an underactuated one. It is redundant because the number of degrees of freedom is much greater than the dimension of the tasks to be performed: around 640 muscles for humans and 30 motors for humanoid robots. It is underactuated because there is no direct actuator allowing the body to move from one place to another place: to do so human and humanoid robots should use their internal degrees of freedom and actuate all their limbs following a periodic process (named bipedal locomotion!).

By considering first that motions are continuous functions from time to space (i.e. trajectories), and second that actions are compositions of motions, actions appear as sequences of trajectories. The images of the trajectories in spaces are named paths. Paths represent geometric traces left by the motions in spaces. The reasoning holds for the real space, the configuration and the control space. Therefore actions appear as continuous simple paths in high dimensional spaces.

A simple path embodies the entire action. It integrates into a single data structure all the complexity of the action. The decomposition of the action into sub-actions (e.g., walk to, grasp, give) appears as the decomposition of the path into sub-paths. Each elementary sub-path is selected among an infinite number of possibilities within some sub-manifolds (e.g., grasp fast or slowly, grasp while bending the legs or not).

All complex cognitive and motor control processes that give rise to an action in the real world are reflected by the structure of paths in the body control space. In this framework, symbols may be defined as sub-manifolds that partition the control space. Such a partition decomposes paths into sub-paths. From this perspective the questions are:

– Motion Segmentation: what are the invariant sub-manifolds that define the structure of a given action?

– Motion Generation: among all the solution paths within a given sub-manifold (i.e. among all the possibilities to solve a given sub-task) what is the underlying law that converges to the selection of a particular motion?

The talk overviews recent results obtained in this framework (including whole body manipulation, locomotion trajectory generation, action recognition) and illustrated from the HRP2-14 humanoid platform.

Jean-Paul Laumond

Cybernics: Fusion of Human, Machine and Information

Robot Suit for the Future

Cybernics is a frontier science centered on cybernetics, mechatronics and informatics, and is a new domain of interdisciplinary academic field of human-assistive technology to support, enhance and expand human’s physical/cognitive functions, which challenges to integrate and harmonize humans and robots (RT: robotics technology) with the basis of information technology (IT) in a functional, organic, and social manner, based on several areas of science and technology such as neuroscience, physiology, robotics, computer science, medicine, behavioral science, ethics, safety engineering, psychology, cognitive science and social science. A pioneering achievement is Robot Suit HAL (Hybrid Assistive Limbs), which is the world’s first cyborg type robot that supports, enhances and strengthens the physical motion of human with the interaction between human and robot by detecting the weak bioelectrical signal through the body from the brain, which generates the nerve signal to control the musculoskeletal system. In this talk, I will deliver the outline of Cybernics and mention about clinical applications for stroke patients, SCI patients, and severe incurable disease such as Neuro-Muscular disease. And I will introduce some remarkable works including new applications of HAL and Vital Sensing System based on Cybernics technologies.

Yoshiyuki Sankai

If Abstraction Is the Answer, What Is the Question? — Reasoning for Everyday Manipulation Tasks

In recent years we have seen tremendous advances in the mechatronic, sensing and computational infrastructure of robots, enabling them to act faster, stronger and more accurately than humans do. Yet, when it comes to accomplishing manipulation tasks in everyday settings, robots are still far away from reaching the sophistication and performance of humans. A key component of the “action intelligence” needed for reaching such a level of sophistication and performance are robot control systems that can take vague action descriptions and automatically infer how they are appropriately executed in a given task and environment context.

Artificial Intelligence (AI) is the research discipline that has studied such reasoning problems for more than fifty years. Researchers in AI have investigated naive physics reasoning, temporal projection, reasoning about action and change, action planning, spatial reasoning, to name only a few. Unfortunately, the proposed methods have not yet achieved their desired impact on autonomous robot control. We believe that one of the reasons is that most AI researchers consider perception and action to be the mere input and output of symbolic reasoning. In contrast, some researchers in cognitive psychology suggest a “simulation model” for reasoning through actions. In their view predicting the consequences of actions is very similar to executing the actions without causing physical effects — perception and action get simulated at a very fine-grained feedback loop.

In this talk I will present reasoning techniques of an autonomous robot control system that are inspired by the “simulation model” for reasoning through actions. These techniques use perception and motor control mechanisms and simulations thereof not only as input and output but more importantly also as resources for symbolic reasoning. I will show, using an autonomous robot making pancakes as an example, that such techniques reason about actions more realistically and thereby enable the robot to improve its performance.

Michael Beetz

Learning and Behavior

Towards Partners Profiling in Human Robot Interaction Contexts

Individuality is one of the most important qualities of humans. Social robots should be able to model the individuality of the human partners and to modify their behaviours accordingly.This paper proposes a profiling system for social robots to be able to learn the individuality of human partners in social contexts. Profiles are expressed in terms of of identities and preferences bound together. In particular, people’s identity is captured by the use of facial features, while preferences are extracted from the discussion between the partners. Both are bound using an Hebb network. Experiments show the feasibility and the performances of the approach presented.

Salvatore M. Anzalone, Yuichiro Yoshikawa, Hiroshi Ishiguro, Emanuele Menegatti, Enrico Pagello, Rosario Sorbello

Motivation-Based Autonomous Behavior Control of Robotic Computer

Successful development of a robotic computer as a mediator in smart environments requires providing a certain level of behavior autonomy to the robot and a capability to adapt its behavior in long-term interaction with the users. We attempt to identify core autonomy-related functionalities and describe the design and implementation of an autonomous behavior control subsystem that provides them. The Motivation Module is essential for providing a balance between the robot’s autonomy and our ability to influence its behavior development in a long term. We present the results of two test scenarios illustrating basic use of the newly provided functionality.

Blagovest Vladimirov, Hyun Kim, Namshik Park

An Evaluation Method for Smart Variable Space in Living Space

The methods which improve space usage efficiency is important especially for city lives. Development of high-rise buildings and underground is one of the methods. However, those developments just lay out the spaces which are expanded in plane into vertical. Physical and monetary limitations are the problem. Therefore, the spaces which can change its functions easily/automatically depending on the situations are necessary instead of stacking of single function spaces. So far, we have proposed Smart Variable Space which realizes various functional spaces by changing its Spatial Configuration Modules dynamically. In this research, the simulated environment for Smart Variable Space was developed by using Virtual Robot Experimentation Platform. In order to clarify the efficacy of Smart Variable Space, a new evaluation index was proposed, and then, the efficacy of Smart Variable Space in living space was assessed by comparison with the conventional housing models.

Kazuyoshi Wada, Keisuke Takayama, Yusuke Suganuma, Toshihiko Suzuki

Modeling Robot Behavior with CCL

This paper presents the use of a

Concurrent Communicating Lists (CCL)

library in robot behavior modeling.

CCL

provides several software components, which allow the model to be built, simulated and formally verified. Due to the integration with the

Robust

library the

CCL

models can be deployed and executed on the actual hardware platforms. Besides the modeling robot behavior, the work also addresses the problem of modeling a robots environment.

The

CCL

models can be verified either formally or by simulation. Since the use of formal methods is always associated with the state explosion problem, the work provides practical guidelines on how to deal with this problem using

CCL

.

Konrad Kułakowski, Tomasz Szmuc

Visual-Trace Simulation of Concurrent Finite-State Machines for Validation and Model-Checking of Complex Behaviour

Simulation of models that specify behaviour of software in robots, embedded systems, and safety critical systems is crucial to ensure correctness. This is particularly important in conjunction with model-driven development, which is highly prevalent due to its numerous benefits. We use vectors of finite-state machines (FSMs) as our modelling tool. Our FSMs can have their transitions labeled by expressions of a common sense logic, and they are more expressive than other modelling approaches (such as Behavior Trees, Petri nets, or plain FSMs). We interpret the models using the same round-robin scheduler which is integrated into the simulator. Execution on a platform is exactly the same as in the simulator (where sensors and actuators are masqueraded by proxies) and coincides with the generator of the Kripke structure for formal model-checking. In three ubiquitous case studies we show that our simulation discovers issues where those models were incomplete, ambiguous, or incorrect. This further illustrates that simulation and monitoring need to complement formal verification.

Robert Coleman, Vladimir Estivill-Castro, René Hexel, Carl Lusty

Modeling of Robots

Fast Dynamic Simulation of Highly Articulated Robots with Contact via Θ(n 2) Time Dense Generalized Inertia Matrix Inversion

The generalized inertia matrix and its inverse are used extensively in robotics applications. While construction of the inertia matrix requires Θ(

n

2

) time, inverting it traditionally employs algorithms running in time

O

(

n

3

). We describe an algorithm that reduces the asymptotic time complexity of this operation to the theoretical minimum: Θ(

n

2

). We also present simple modifications that reduce the number of arithmetic operations (and thereby the running time). We compare our approach against fast Cholesky factorization both theoretically (using number of arithmetic operations) and empirically (using running times). We demonstrate our method to dynamically simulate a highly articulated robot undergoing contact, yielding an order of magnitude decrease in running time over existing methods.

Evan Drumwright

A Differential-Algebraic Multistate Friction Model

Fidelity with friction properties and easiness of implementation are both important aspects for friction modeling. Some empirically motivated models can be implemented easily due to their simple expression and small number of parameters, but they cannot capture faithfully the main properties of friction. Some physically motivated models give close agreement with the friction properties, but they can be too complex for some applications. This paper proposes a differential-algebraic multistate friction model that possesses easiness of implementation and adjustment, a relatively small number of parameters and a compact formulation. Moreover, it captures all standard properties of well-established friction models.

Xiaogang Xiong, Ryo Kikuuwe, Motoji Yamamoto

Simulation of Flexible Objects in Robotics

In this paper, we present what appears to be the first simulation model for grasping of flexible bodies based on the three-dimensional elastic constitutive relations and Newton’s Second Law for solids known as the Navier-Cauchy equations. We give an overview of the most important equations for strain, stress, and elasticity tensors based on which we outline the format of the Navier-Cauchy equations of motion in the general anisotropic case. We then specifically study the equations for homogeneous isotropic bodies. We formulate a numerical scheme based on finite differences for solving the equations. Finally, we present preliminary experimental work and outline future directions.

Andreas Rune Fugl, Henrik Gordon Petersen, Morten Willatzen

Continuous Integration for Iterative Validation of Simulated Robot Models

Simulated environments often provide the first, and are usually the most frequent, test environment for robotic systems, primarily due to their cost and safety advantages. Unfortunately, changing aspects of both, the simulation and the real robot, as well as actuator control algorithms are often not taken into account when relying on simulation results. In this paper we present a continuous integration approach to verify simulated robot models in an integrated and frequent manner, comprising a simulated and a real robot for comparison. The central aspect of our concept is to iteratively assess the fidelity of simulated robot models. In an exemplary case study we distilled a first set of requirements and metrics, which can be used by developers to verify their algorithms and to automatically detect further system changes.

Florian Lier, Simon Schulz, Ingo Lütkebohle

Software Modeling and Architecture

Software Abstractions for Simulation and Control of a Continuum Robot

The Bionic Handling Assistant is a new continuum robot which is manufactured in a rapid-prototyping procedure out of elastic polyamide. Its mechanical flexibility and low weight provide an enormous potential for physical human robot interaction. Yet, the elasticity and parallel continuum actuation design challenge standard approaches to deal with a robot from a control, simulation, and software modeling perspective. We investigate how the software abstractions of the existing Robot Control Interface (RCI) and the Compliant Control Architecture (CCA) can deal with this platform from a software modeling and software architectural perspective. We focus on three different challenges: the first challenge is to enable reasonable and hierarchical

semantic abstractions

of the robot. The second challenge is to develop

hardware I/O abstractions

for the prototypical and heterogeneous technical setup. The third challenge is to realize this in a flexible and reusable manner. We evaluate our approaches to the above challenges in a practical scenario in which the robot is controlled either in simulation or on the real robot.

Arne Nordmann, Matthias Rolf, Sebastian Wrede

A Visual Modeling Language for RDIS and ROS Nodes Using AToM3

In robotics we are often faced with the problem of repeatedly writing robot drivers for the same devices, but different robot frameworks. In an effort to counter this, a domain specific language for generating robot drivers was developed. However, descriptions tend to get verbose fast and the adopted syntax was difficult for programmers. This paper outlines an attempt to shift away from a textual syntax and toward a visual syntax, instead relying on the textual syntax to communicate the model to other tools. In addition, a formalism for defining ROS nodes is presented and a model transformation for mapping RDIS messages to ROS messages and vice-versa is created.

Paul Kilgo, Eugene Syriani, Monica Anderson

PRACSYS: An Extensible Architecture for Composing Motion Controllers and Planners

This paper describes a software infrastructure for developing controllers and planners for robotic systems, referred here as

PRACSYS

. At the core of the software is the abstraction of a dynamical system, which, given a control, propagates its state forward in time. The platform simplifies the development of new controllers and planners and provides an extensible framework that allows complex interactions between one or many controllers, as well as motion planners. For instance, it is possible to compose many control layers over a physical system, to define multi-agent controllers that operate over many systems, to easily switch between different underlying controllers, and plan over controllers to achieve feedback-based planning. Such capabilities are especially useful for the control of hybrid and cyber-physical systems, which are important in many applications. The software is complementary and builds on top of many existing open-source contributions. It allows the use of different libraries as plugins for various modules, such as collision checking, physics-based simulation, visualization, and planning. This paper describes the overall architecture, explains important features and provides use-cases that evaluate aspects of the infrastructure.

Andrew Kimmel, Andrew Dobson, Zakary Littlefield, Athanasios Krontiris, James Marble, Kostas E. Bekris

RobotML, a Domain-Specific Language to Design, Simulate and Deploy Robotic Applications

A large number of robotic software have been developed but cannot or can hardly interoperate with each other because of their dependencies on specific hardware or software platform is hard-wired into the code. Consequently, robotic software is hard and expensive to develop because there is little opportunity of reuse and because low-level details must be taken into account in early phases. Moreover, robotic experts can hardly develop their application without programming knowledge or the help of programming experts and robotic software is difficult to adapt to hardware or target-platform changes. In this paper we report on the development of

RobotML

, a Robotic Modeling Language that eases the design of robotic applications, their simulation and their deployment to multiple target execution platforms.

Saadia Dhouib, Selma Kchir, Serge Stinckwich, Tewfik Ziadi, Mikal Ziane

A Java vs. C++ Performance Evaluation: A 3D Modeling Benchmark

Along the years robotics software and applications have been typically implemented in compiled languages, such as C and C++, rather than interpreted languages, like Java. This choice has been due to their well-known faster behaviors, which meet the high performance requirements of robotics. Nevertheless, several projects that implement robotics functionality in Java can be found in literature and different experiments conduced by computer scientists have proved that the difference between Java and C++ is not so evident.

In this paper we report our work on quantifying the difference of performance between Java and C++ and we offer a set of data in order to better understand whether the performance of Java allows to consider it a valid alternative for robotics applications or not. We report about the execution time of a Java implementation of an algorithm originally written in C++ and we compare this data with the performance of the original version. Results show that, using the appropriate optimizations, Java is from 1.09 to 1.51 times slower than C++ under Windows and from 1.21 to 1.91 times under Linux.

Luca Gherardi, Davide Brugali, Daniele Comotti

Simulation and Applications

A Comparison of Sampling Strategies for Parameter Estimation of a Robot Simulator

Methods for dealing with the problem of the “reality gap” in evolutionary robotics are described. The focus is on simulator tuning, in which simulator parameters are adjusted in order to more accurately model reality. We investigate sample selection, which is the method of choosing the robot controllers, evaluated in reality, that guide simulator tuning. Six strategies for sample selection are compared on a robot locomotion task. It is found that strategies that select samples that show high fitness in simulation greatly outperform those that do not. One such strategy, which selects the sample that is the expected fittest as well as the most informative (in the sense of producing the most disagreement between potential simulators), results in the creation of a nearly optimal simulator in the first iteration of the simulator tuning algorithm.

Gordon Klaus, Kyrre Glette, Jim Tørresen

A Framework with a Pedestrian Simulator for Deploying Robots into a Real Environment

We describe a simulation framework aimed to develop and test robots before deploying them in a real environment crowded with pedestrians. In order to use mobile robots in the real world, it is necessary to test whether they are able to navigate well, i.e. without causing safety risks to humans. This task is particular difficult due to the complex behavior pedestrians have towards each other and also towards the robot, that can be perceived either as an obstacle to avoid or as an object of interest to approach for curiosity. To overcome this difficulty, our framework involves a pedestrian simulator, based on a collision avoidance model developed to describe low density conditions as those occurring in shopping malls, to test the robot’s navigation capability among pedestrians. Furthermore, we analyzed the behavior of pedestrians towards a robot in a shopping mall to build a human-to-robot interaction model that was introduced in the simulator. Our simulator works as a tool to test the level of safety of robot navigation before deploying it in a real environment. We demonstrate our approach showing how we used the simulator, and how the robot finally navigated in a real environment.

Masahiro Shiomi, Francesco Zanlungo, Kotaro Hayashi, Takayuki Kanda

Simulating Complex Robotic Scenarios with MORSE

MORSE is a robotic simulation software developed by roboticists from several research laboratories. It is a framework to evaluate robotic algorithms and their integration in complex environments, modeled with the Blender 3D real-time engine which brings realistic rendering and physics simulation. The simulations can be specified at various levels of abstraction. This enables researchers to focus on their field of interest, that can range from processing low-level sensor data to the integration of a complete team of robots. After nearly three years of development, MORSE is a mature tool with a large collection of components, that provides many innovative features: software-in-the-loop connectivity, multiple middleware support, configurable components, varying levels of simulation abstraction, distributed implementation for large scale multi-robot simulations and a human avatar that can interact with robots in virtual environments. This paper presents the current state of MORSE, highlighting its unique features in use cases.

Gilberto Echeverria, Séverin Lemaignan, Arnaud Degroote, Simon Lacroix, Michael Karg, Pierrick Koch, Charles Lesire, Serge Stinckwich

Humanoid and Biped Robots

Masters’ Skill Explained by Visualization of Whole-Body Muscle Activity

In this paper, we discuss the computation of human motion dynamics and its analysis of experts’ motion skills. The computation framework of the wire-driven multi-body dynamics previously developed by the authors is applied to the whole body human musculoskeletal model. While capturing time-series of motion data, somatosensory information measured by force plate and EMG sensors simultaneously is used for the dynamics computation. For the dynamics computation, we reduced the computation cost drastically by resolving to the non-linear optimization problem using decomposed gradient computation developed recently. As examples of analysis, we measured and analyzed experts’ motion patterns, such as Tai Chi motion, tap dance and drum playing. In particular, we analyzed the characteristic behavior by the motion of the center of gravity (COG), condition of the ground contact, and muscle activity of the whole body.

Yosuke Ikegami, Ko Ayusawa, Yoshihiko Nakamura

Studying the Effect of Different Optimization Criteria on Humanoid Walking Motions

The generation of stable, efficient and versatile walking motions for humanoid robots is still an open field of research. Several approaches have been implemented on humanoids in the past years, but so far none has led to a walking performance that is anywhere close to humans. This may be caused by limitations of the robotic hardware, but we claim that it is also due to the methods chosen for motion generation which do not fully exploit the capabilities of the hardware. Often, several characteristics of the gait, such as foot placement or step time, are fixed in advance in a suboptimal way for the robot. In this paper we discuss the potential of our optimal control techniques based on dynamical models of the humanoid robot for the generation of improved walking motions. We apply the method to a 3D dynamic model of the humanoid robot HRP-2 with 36 DOF and 30 actuators. Robot specific stability constraints (such as ZMP constraints) can be taken into account in the optimization. We present results for five different objective functions, and evaluate the influence of free foot placement and a relaxation of ZMP constraints.

Kai Henning Koch, Katja Daniela Mombaur, Philipp Souères

Modeling and Simulating Compliant Movements in a Musculoskeletal Bipedal Robot

This paper describes the modeling and the simulation of a novel Elastic Bipedal Robot based on Human Musculoskeletal modeling. The geometrical organization of the robot artificial muscles is based on the organization of human muscles. In this paper we study how the robot active and passive elastic actuation structures develop force during selected motor tasks, and how we can model the contact between feet and ground. We then compare the robot dynamics to that of the human during the same motor tasks. The motivation behind this study is to reduce the development time by using a simulation environment for the purpose of developing a bipedal robot that takes advantage of the mechanisms underlying the human musculoskeletal dynamics for the generation of natural movement.

Roberto Bortoletto, Massimo Sartori, Fuben He, Enrico Pagello

Simulation and Experimental Evaluation of the Contribution of Biarticular Gastrocnemius Structure to Joint Synchronization in Human-Inspired Three-Segmented Elastic Legs

The humanoid robot BioBiped2 is powered by series elastic actuators (SEA) at the leg joints. As motivated by the human muscle architecture comprising monoarticular and biarticular muscles, the SEA at joint level are supported by elastic elements spanning two joints. In this study we demonstrate in simulation and in robot experiments, to what extend synchronous joint operation can be enhanced by introducing elastic biarticular structures in the leg, reducing the risk of over-extending individual joints.

Dorian Scholz, Christophe Maufroy, Stefan Kurowski, Katayon Radkhah, Oskar von Stryk, André Seyfarth

Mobile Robots

Graph Optimization with Unstructured Covariance: Fast, Accurate, Linear Approximation

This manuscript addresses the problem of optimization- based Simultaneous Localization and Mapping (SLAM), which is of concern when a robot, traveling in an unknown environment, has to build a world model, exploiting sensor measurements. Although the optimization problem underlying SLAM is nonlinear and nonconvex, related work showed that it is possible to compute an accurate linear approximation of the optimal solution for the case in which measurement covariance matrices have a block diagonal structure. In this paper we relax this hypothesis on the structure of measurement covariance and we propose a linear approximation that can deal with the general unstructured case. After presenting our theoretical derivation, we report an experimental evaluation of the proposed technique. The outcome confirms that the technique has remarkable advantages over state-of-the-art approaches and it is a promising solution for large-scale mapping.

Luca Carlone, Jingchun Yin, Stefano Rosa, Zehui Yuan

Mobile Robot SLAM Interacting with Networked Small Intelligent Sensors Distributed in Indoor Environments

SLAM is a method of map building and self-position estimation for robot navigation. However, map building error is especially appeared in loop closing points when the mobile robot moves around loop trajectories. In this study, more accurate mobile robot SLAM is considered in intelligent space [1] where many sensors are distributed.

An intelligent space is constructed with various types of distributed sensors including networked laser range sensors. Laser sensor on a mobile robot and environment sensors share sensor information each other in intelligent space. Maps and self-positions of the mobile robot are estimated using geometrical relationships between the mobile robot and sensors in intelligent space. However, geometrical calibration of distributed sensors under the unified world coordinates is required for construction of the intelligent space. When many sensors are distributed in wide area, it generally becomes complicated tasks to calibrate all sensors. In order to solve these problems, we consider extend SLAM algorithm. In this study, a new method of SLAM, which uses distributed sensors fixed in the intelligent space, is introduced. This method aims to achieve precision SLAM and position estimation of networked laser range sensors in the intelligent space.

Fumitaka Hashikawa, Kazuyuki Morioka, Noriaki Ando

Manipulation

Computing 2D Robot Workspace in Parallel with CUDA

Workspace analysis is one of the most essential problems in robotics, but also has the possibility of being very tricky in complex cases. As the number of degrees of freedom increases, the complexity of the problem grows exponentially in some solutions. One possibility is to develop solutions which approximate the workspace for speedup, but this paper explores the possibility of using graphical processing units to parallelize and speed up a forward kinematics-based solution. Particular real-time applications are discussed. It presents a formal analysis of a simple 2D problem, a solution, and the results of an experiment using the solution.

Paul Kilgo, Brandon Dixon, Monica Anderson

Acquisition of Object Pose from Barcode for Robot Manipulation

General, robots obtain poses of target objects by matching the images of the observed objects with data in database. However, the process of matching images costs so long time that robot’s action become slow. In order to shorten response time for robots searching target objects, we propose a method for robots to obtain information of poses of observed objects by calculating corner points of barcodes on the objects. Since information in a barcode is less than the one in an image, the method can help robot rapidly obtain the information of its target objects. Furthermore, in order to reuse the method in other robot systems, we create a RT-Component(RTC) to realize the method.

Yuexing Han, Yasushi Sumi, Yoshio Matsumoto, Noriaki Ando

WorkCellSimulator: A 3D Simulator for Intelligent Manufacturing

This paper presents WorkCellSimulator, a software platform that allows to manage an environment for the simulation of robot tasks. It uses the most advanced artificial intelligence algorithms in order to define the production process, by controlling one or more robot manipulators and machineries present in the work cell. The main goal of this software is to assist the user in defining customized production processes which involve specific automated cells. It has been developed by IT+Robotics, a spin-off company of the University of Padua, founded in 2005 from the collaboration between young researchers in the field of Robotics and a group of professors from the Department of Information Engineering, University of Padua.

Stefano Tonello, Guido Piero Zanetti, Matteo Finotto, Roberto Bortoletto, Elisa Tosello, Emanuele Menegatti

Tools and Middleware

A Meta-model and Toolchain for Improved Interoperability of Robotic Frameworks

The emerging availability of high-quality software repositories for robotics promises to speed up the construction process of robotic systems through systematic reuse of software components. However, to reuse components without modification, compatibility at the interface level needs to be created, which is particularly hard if components were implemented in different robotic frameworks. In this paper we propose an approach using model-based techniques for improving component reusability. We specifically address data type compatibility in a structured way through the development of a generic meta-model capable of representing data types from different frameworks and their relations. Based on this model a code generator emits serialization code which makes it possible to seamlessly reuse the existing data types of different frameworks. The application of this approach is exemplified by connecting the YARP-based iCub simulation with a component architecture using a current robotics middleware. Based on our experiences we describe requirements on robotics frameworks to further increase the level of interoperability between available components.

Johannes Wienke, Arne Nordmann, Sebastian Wrede

Integrated Software Development for Embedded Robotic Systems

In the recent years, improvements in robotic hardware have not been matched by advancements in robotic software and the gap between those two areas has been widening. To cope with the increasing complexity of novel robotic embedded systems an integrated and continuous software development process is required supporting different development activities and stages being integrated into an overall development methodology, supported by libraries, elaborated tools and toolchains. For an efficient development of robotic systems a seamless integration between different activities and stages is required. In the domain of automotive systems, such an overall development methodology, consisting of different development activities/stages and supported by elaborated libraries, tools and toolchains, already exists. In this paper, we show how to adapt an existing methodology for the development of automotive embedded systems for being applicable on robotic systems.

Sebastian Wätzoldt, Stefan Neumann, Falk Benke, Holger Giese

Combining IEC 61499 Model-Based Design with Component-Based Architecture for Robotics

The model-driven approach is an increasingly popular trend in software design. It provides many benefits in terms of system design, reusability, and automatic code generation. In the industrial automation domain, the IEC 61499 standard is a recent initiative that adopts this approach. It offers an open, platform-independent framework for designing distributed control systems, whereby the interface and behaviour of a component is described using a function block. On the other hand, in the robotics domain, the Robotic Technology Component specification proposes a framework that allows software components to be easily integrated in a robotic system. The focus of that specification is not so much on the definition of each component’s internal behaviour, but rather on the management and interaction of those components. The combination of both these standards offers a comprehensive solution for designing robotic software components in a model-driven approach. This paper describes a tool-chain for doing so, and illustrates its viability through an example.

Li Hsien Yoong, Zeeshan E. Bhatti, Partha S. Roop

A Reuse-Oriented Development Process for Component-Based Robotic Systems

State of the art in robot software development mostly relies on class library reuse and only to a limited extent to component-based design. In the BRICS project we have defined a software development process that is based on the two most recent and promising approaches to software reuse, i.e. Software Product Line (SPL) and Model-Driven Engineering (MDE). The aim of this paper is to illustrate the whole software development process that we have defined for developing flexible and reusable component-based robotics libraries, to exemplify it with the case study of robust navigation functionality, and to present the software tools that we have developed for supporting the proposed process.

Davide Brugali, Luca Gherardi, A. Biziak, Andrea Luzzana, Alexey Zakharov

UAV Simulation

SwarmSimX: Real-Time Simulation Environment for Multi-robot Systems

In this paper we present a novel simulation environment called SwarmSimX with the ability to simulate dozens of robots in a realistic 3D environment. The software architecture of SwarmSimX allows new robots, sensors, and other libraries to be loaded at runtime, extending the functionality of the simulation environment significantly. In addition, SwarmSimX allows an easy exchange of the underlying libraries used for the visual and physical simulation to incorporate different libraries (e.g., improved or future versions). A major feature is also the possibility to perform the whole simulation in real-time allowing for human-in-the-loop or hardware-in-the-loop scenarios. SwarmSimX has been already employed in several works presenting haptic shared control of multiple mobile robots (e.g., quadrotor UAVs). Additionally, we present here two validation tests showing the physical fidelity and the real-time performance of SwarmSimX. For the tests we used NVIDIA

®

PhysX

®

and Ogre3D as physics and rendering libraries, respectively.

Johannes Lächele, Antonio Franchi, Heinrich H. Bülthoff, Paolo Robuffo Giordano

Evaluating the Effectiveness of Mixed Reality Simulations for Developing UAV Systems

The development cycle of an Unmanned Aerial Vehicle (UAV) system can be long and challenging. Mixed Reality (MR) simulations can reduce cost, duration and risk of the development process by enabling the replacement of expensive, dangerous, or not yet fully developed components with virtual counterparts. However, there has been little validation of such hybrid simulation methods in practical robot applications. This paper evaluates the use of MR simulations for prototyping a UAV system to be deployed for a dairy farming monitoring task. We show that by augmenting the robot’s sensing with a virtual moving cow using an extensible Augmented Reality (AR) tracking technique, MR simulations could help to provide efficient testing and identify improvements to the UAV controller. User study findings reveal the importance of both virtual and MR simulations to robot development, with MR simulations helping developers transition to development in a more physical environment.

Ian Yen-Hung Chen, Bruce MacDonald, Burkhard Wünsche

Comprehensive Simulation of Quadrotor UAVs Using ROS and Gazebo

Quadrotor UAVs have successfully been used both in research and for commercial applications in recent years and there has been significant progress in the design of robust control software and hardware. Nevertheless, testing of prototype UAV systems still means risk of damage due to failures. Motivated by this, a system for the comprehensive simulation of quadrotor UAVs is presented in this paper. Unlike existing solutions, the presented system is integrated with ROS and the Gazebo simulator. This comprehensive approach allows simultaneous simulation of diverse aspects such as flight dynamics, onboard sensors like IMUs, external imaging sensors and complex environments. The dynamics model of the quadrotor has been parameterized using wind tunnel tests and validated by a comparison of simulated and real flight data. The applicability for simulation of complex UAV systems is demonstrated using LIDAR-based and visual SLAM approaches available as open source software.

Johannes Meyer, Alexander Sendobry, Stefan Kohlbrecher, Uwe Klingauf, Oskar von Stryk

Backmatter

Weitere Informationen