连续第四年ROSCon门票完全售罄，参加人数超过500人。感谢47个赞助商6，数量再创新高， 为会议的发展提供资金支持！
此外，还发布了几乎所有讲座的幻灯片。因此，如果想获得链接或其他信息，现在也可以这样做。幻灯片与来自程序册（会议日程安排）60 的视频链接。
如果提出Lightning Talk并希望提供更多信息，或者如果无法获得简要报告时段，在ROS项目类别1，已发布的ROS包，请将其发送到“ 常规”类别。这样，更多的社区将了解项目和功能包贡献。
- ROSCon 2018组委会
Thank you again to our Platinum Sponsor, Erle, and to all of our Gold Sponsors: Amazon, Apple, ARM, Clearpath, Eprosima, Fetch Robotics, Google, Intel, Locus, Microsoft, ROBOTIS, SICK, Tier IV, Toyota Research Institute, Universal Robots.
----2012~2017年会议资料链接----
- The recordings and slides have been posted in the program
- Archives from 2012, 2013, 2014, 2015, 2016, and 2017 are available.
----
Day 1, September 29th
| Time | Presenter(s) | Title | Abstract | Resources |
|---|---|---|---|---|
| 7:30 | Everyone | Registration open | Please arrive early to allow time to collect your badge and conference bag before the presentations start. We expect there to be a queue for registration on the first day. | |
| 9:00 | Brian Gerkey (Open Robotics) | Opening Remarks | VideoSlides | |
| 9:10 | Andrew Symington (NASA, Stinger Ghaffarian Technologies & KBRWyle) | Astrobee: ROS-based Flight Software for a Free-flying Robot in Microgravity | Astrobee is a free-flying robot designed by NASA to operate alongside astronauts inside the International Space Station, where it will carry out scientific and surveying tasks in microgravity. The robot can autonomously mate with a docking station to recharge, as well as perch to existing ISS handrails using a three degrees of freedom arm. Its open source flight software stack is built on ROS Kinetic, uses a delay-tolerant DDS bridge for space-to-ground communication, and is accompanied by a Gazebo simulator that enables researchers to develop and test behavioral algorithms. This presentation covers the the software architecture, challenges faced during the development process, facilities for testing prototype hardware, and broad lessons we have learned over the last three years. | VideoSlides |
| 9:40 | Ian Chen (Open Robotics), Mark Allan (KBR, Inc.) | Gazebo renders the moon | Over the past couple of years, Open Robotics collaborated with NASA Ames Intelligent Robotics Group to develop the Resource Prospector lunar driving simulator. In particular, a strong focus was placed on generating high quality visual images from the cameras on the simulated rover. This resulted in in several improvements made to Gazebo, including support of high resolution digital elevation maps, improved shadows, integration of custom material shaders, and a new lens flare plugin. This presentation will show the resulting lunar terrain environment created in Gazebo and discuss some of the challenges in modeling the environment. | VideoSlides |
| 10:00 | Musa Morena Marcusso Manhães (Corporate Sector for Research and Advance Engineering - Robert Bosch GmbH) | Unmanned Underwater Vehicle Simulator: Enabling the Simulation of Multi-Robot Underwater Missions with Gazebo | Since 2016, the Unmanned Underwater Vehicle Simulator has been developed as an open-source project that extends Gazebo to enable the simulation of multi-robot underwater vehicles in the scope of the EU project SWARMs. This presentation will show the key features in the package, such as implementations of plugins for computation hydrodynamic and hydrostatic forces, and typical underwater actuator models, and demonstrations on use-cases derived from the SWARMs project involving multi-robot plume tracking and seabed mapping missions. | VideoSlides |
| 10:20 | Nicolo Valigi (Cruise Automation) | Lessons learned building a self-driving car on ROS | Cruise Automation’s self driving car runs on top on ROS. This talk will share some of the lessons we learned while scaling up the ROS stack to a very complex Robotics problem and 500+ engineers. We will talk about performance, reliability, code organization and health, and the ways we have found ROS to excel or fall short. | VideoSlides |
| 10:30 | Everyone | Coffee Break | Sponsored by Honeywell Intelligrated | |
| 11:00 | Mikael Arguedas (Open Robotics), Karsten Knese (Robert Bosch LLC) | Hands-on ROS 2: A Walkthrough | In this talk we show the current state of ROS 2 features through hands-on demonstrations. We guide the developer through a set of features and functionalities of ROS 2 beginning with creating a simple “hello world” package and consecutively increasing functionality. We show how to launch multiple nodes, how lifecycle nodes can be used to bootstrap a complete system, through to how the ROS 2 security features can be utilized to secure applications. This talk is meant to highlight the newly available features and tools of ROS 2 and aims to incline developers to start working with it. | VideoSlides |
| 11:30 | Christopher Ho (Apex.AI, Inc), Sumanth Nirmal (Apex.AI, Inc), Dejan Pangercic (Apex.AI, Inc), Shinpei Kato (Tier IV, Inc), Jan Becker (Apex.AI, Inc) | ROS 2 on Autonomous Driving Vehicles | Over the past 10 years, ROS 1 has proven itself to be the framework of choice for prototyping and developing large robotic applications. Many limitations preventing ROS 1 from being used in production applications have been discovered over the years, and after significant prototyping, ROS 2 makes great headway in making ROS 2 suitable for production. Autonomous driving is the next great technology waiting to be realized and transform our society. A full autonomous driving stack is inarguably a large robotic system, and consequently ROS 2 is the right framework upon which to develop a full autonomous driving stack. To prove this, we have developed a part of the autonomous driving stack based on ROS 2. | VideoSlides |
| 11:50 | Andreas Holzner (TNG Technology Consulting GmbH) | RViz – The tale of a migration to ROS 2.0 | RViz is now available for ROS 2, including most of its features. This talk will look back at the migration of RViz from ROS 1 to ROS 2, and present the main challenges and changes performed during the migration. It will focus on a new package, rviz_visual_testing_framework, that makes writing automated UI tests - including the 3D rendering part of RViz – possible. | VideoSlides |
| 12:10 | William Woodall (Open Robotics) | Launch for ROS 2 | Over the past half year we have been working on the new launch system for ROS 2 based on roslaunch from ROS 1. The goal of this presentation is to summarize the state of the design document for launch, describe the current state of the reference implementation, and dive into the rationale behind some of the design decisions. The attendee should come away with a deeper understanding of the differences between roslaunch from ROS 1 and the new ros2 launchtool. We’ll also have some short demonstrations to show how these differences might be useful to everyday users. |
VideoSlides |
| 12:20 | Deanna Hood (Open Robotics), Mikael Arguedas (Open Robotics) | Getting involved in ROS 2 development | Let us introduce you to the development process for ROS and how you can get involved in the ROS 2 project. We will cover a number of ways to participate, suiting a variety of profiles, including ROS 2 users looking to make their first open source contribution, organizations interested in supporting the ROS 2 project, and even experienced ROS 1 users with no time for software contributions. | VideoSlides |
| 12:30 | Everyone | Lunch | Sponsored by ADLINK | |
| 13:50 | Steve Macenski (Simbe Robotics) | Use of the Spatio-Temporal Voxel Layer | We introduce the Spatio-Temporal Voxel Grid and accompanying costmap_2d plugin, Spatio-Temporal Voxel Layer (STVL) , for the ROS Navigation Stack. It is implemented as a drop-in replacement for the current VoxelLayer plugin. STVL brings several improvements over the existing VoxelLayer including improved voxel grid representation, Spatio-Temporal based voxel clearing rather than Bresenham’s-based clearing, and lower computational burdens to run on low-powered computers effectively. This package was built for perception in retail, warehouse, and outdoor robot applications, though effective out of those scopes. This talk will go over key points of STVL, demonstrations in production environments, and how to enable it in your application. | VideoSlides |
| 14:10 | Margarida Costa Faria (Center for Advanced Aerospace Technologies), Ivan Maza (Robotics, Vision and Control Group, University of Seville), Antidio Viguria (Center for Advanced Aerospace Technologies) | Lazy_theta_star - a deterministic 3D path planner | Generating paths in large, outdoor, 3D scenarios, online is crucial to autonomous vehicles in various environments like air or underwater. Deterministic algorithms can give more guarantees on the action taken in a given situation, a feature highly important in standardization and verification. The path_planning package offers the implementation of Lazy Theta *, extended for 3D and implemented over octomap. This combination addresses all of the mentioned concerns. The talk will focus on how to set up a service that responds with 3D obstacle-avoiding paths, in what scenarios to use it, and a brief contextualization of the algorithm. | VideoSlides |
| 14:20 | David V Lu!! (Locus Robotics) | Planning to Plan: New Flexible Navigation Interfaces | The navigation stack is one of the core components of the ROS Ecosystem. This talk will present lessons learned from maintaining the navigation stack over the past five years, and discuss a new generation of interfaces for increased functionality. This talk will cover: the new nav_core2 interfaces, with a new Costmapinterface; the locomotor package which replaces move_base and demonstrates new functionality; and ROS 2.0 + Navigation. |
VideoSlides |
| 14:40 | Gerardo Pardo-Castellote (Real-Time Innovations), Ruffin White (University of California, San Diego) | Leveraging DDS Security in ROS2 | The ROS2 Middleware Interface (rmw) uses OMG Data-Distribution Service (DDS) standard middleware and the DDS-RTPS protocol. One key aspect of DDS is the support for secure communications defined by the DDS-Security standard. This presentation describes the main characteristics of DDS-Security: How applications are identified and authenticated, access control mechanisms, and dataflow protection options (e.g., encryption, authentication, origin authentication). We will describe interesting scenarios and how they would impact the deployment of secure ROS2 systems that leverage DDS-Security. We also include benchmarks showing the expected performance impact resulting from each of these choices. | VideoSlides |
| 15:10 | Filipe Rinaldi (Arm), Matt Spencer (Arm) | Arm DDS Security library: Adding secure security to ROS2 | In this session we will present the Arm DDS Security library project: an open source library that isolates the DDS security operations and key management under a Trusted Execution Environment (TEE). This approach uses Arm security hardware - TrustZone - to secure communication between discrete components w ithin the fabric of autonomous machines - Automotive, Robotics, Drones etc. - whilst keeping secure assets (e.g. keys) protected. | VideoSlides |
| 15:20 | Everyone | Lightning Talks I | 1 - CARMA - Nuclear Inspection Robot (video) (slides) 2 - Gaitech BCI: Bringing Brain Computer Interfacing to ROS (video) (slides) 3 - Rosbag for ROS 2 - development status (video) (slides) 4 - Mixed Real-Time Criticality with ROS2 - the Callback-group-level Executor (video) (slides) 5 - gita : Making a Consumer Robot (video) (slides) 6 - ROSCon JP (video) (slides) 7 - Robot Recorder (video) (slides) 8 - ROS and Retail Robotics (video) (slides) 9 - Service Robotics Modules by Pilz (video) (slides) 10 - SMACH SCXML Interpreter (video) (slides) 11 - XEL Network : modular H/W ecosystem using ROS2 (video) (slides) 12 - Feeling by seeing - RoVi Robot Vision (video) (slides) 13 - HRIM, The Hardware Robot Information Model (video) (slides) 14 - ROS2 on WebOS: Web-app enabled robots (video) 15 - Autonomous Simulator for Baidu Apollo (video) |
|
| 16:05 | Everyone | Coffee Break | Sponsored by Honeywell Intelligrated | |
| 16:35 | Jerry Towler (Southwest Research Institute) | ROS-Military: Progress and Promise | The United States military has invested in robotics for decades for applications from long-range reconnaissance and surveillance to directly interacting with and supporting soldiers on the ground. However, over much of that time, its programmatic and technical approaches to robotics have remained traditional and monolithic. Supported by robotics development companies, academic researchers, the U.S. Government, and the Open Source Robotics Foundation itself, ROS-Military is a multi-faceted approach to effectively bringing the collaborative success of the ROS ecosystem to U.S. military robotics and, wherever possible, contributing its achievements back to the global ROS community. | Video |
| 16:55 | Toffee Albina (Toyota Research Institute), Chris Lalancette (Open Robotics) | ROS2: Supercharging the Jaguar4x4 | In this talk we’ll show the viability of using ROS2 for real-world robotics. The talk will focus on the process of creating a ROS2 platform for the DrRobot Jaguar4x4 robot (with the manipulator). We’ll discuss the open-source work done to get the robot base and manipulator working, as well as some of the challenges we encountered while doing this work. We’ll also discuss various ways of getting navigation to work in ROS2 on this base. | VideoSlides |
| 17:15 | Andreas Pasternak (Apex.AI, Inc), Dejan Pangercic (Apex.AI, Inc) | Performance Test - A Tool for Communication Middleware Performance Measuring | Verifying that your communication middleware and hardware platform satisfies throughput and latency requirements can be challenging. Often, several middleware options are available, each with a rich set of configuration options and testing them in a consistent way is desirable. We have released performance_test, a C++ tool for testing different communication middlewares. An exhaustive logging to file and accompanying python tools to parse and plot the data make it easy to verify that a system can satisfy the desired communication performance requirements. Furthermore, its plugin based architecture makes it easy to test currently unsupported communication middlewares. Currently supported middlewares are ROS 2 through rmw layer and thus all underlying DDS implementations. Further, FastRTPS and Connext Micro DDS are supported directly. In this session we will talk about the difference a real time operating system makes regarding latency stability. Further, we will demonstrate how to run tests and visualize the data. At last, we will give an overview on how to create plugins for other communication middlewares. | VideoSlides |
| 17:25 | Adam Alami (IT University of Copenhagen), Zhoulai Fu (IT University of Copenhagen) | ROSIN Quality Assurance Initiatives for ROS | ROSIN is an EU Horizon 2020 research project aiming at improving the quality of ROS and ROS-Industrial. We will inform on our involvement in two major Quality Assurance (QA) initiatives of the project: (1) Community engagement: We have established a working group of community members for implementing QA initiatives. We will report on the activities of the group. (2) Tool development: We have been developing an automated method for blackbox testing ROS components based on fuzzing. Fuzzing is the technique that detected Heartbleed. We will demonstrate how users can use our tool and integrate it into the ROS development cycle. | Video 1Video 2Slides 1Slides 2 |
| 17:45 | Michael Grupp (Magazino GmbH), Andre Gaschler (Google Germany GmbH), Mirko Ferrati (Magazino GmbH), Christoph Schütte (Google Germany GmbH) | Cloud-based Mapping and Localization in Dynamic Warehouse Environments | Reliable, life-long mapping and localization is an essential component for mobile robotics in continuously changing warehouse environments. We present a system based on Cartographer in which robots run finite-history SLAM for low-latency localization and continuously stream local map updates to a cloud service. The cloud component assembles and optimizes a globally consistent pose graph out of the streaming data of the agents. Magazino piloted cloud-based Cartographer in a customer warehouse using its fleet of mobile picking robots. By sharing the local map changes among each other, the robots were able to maintain their localization accuracy while dealing with dynamic environments. | Video |
| 18:05 | Everyone | Group Photo | ||
| 18:10 | Everyone | Reception | Sponsored by Google | |
| 20:05 | Fin |
Day 2, September 30th
| Time | Presenter(s) | Title | Abstract | Resources |
|---|---|---|---|---|
| 8:30 | Everyone | Registration open | ||
| 9:00 | Tomoya Fujita (Sony Corporation) | aibo development using ROS | On January 11, 2018, Sony Corporation released aibo (https://aibo.sony.jp/). aibo that is back on market beyond the time of 12 years constructed via robotics framework named ROS. In this presentation, we introduce examples of development in aibo from the point of view of ROS, starting with introduction of aibo, architecture, embedded technology, real-time optimization, robot development environment, simulation etc. | VideoSlides |
| 9:30 | Marshall Massengill (The Zebracorns, VMware), Kevin Jaget (The Zebracorns, Nvidia), Megan Wenzinger (The Zebracorns, IBM), Niall Mullane (The Zebracorns, NCSSM), Anja Sheppard (The Zebracorns, UT Dallas), Olivia Fugikawa (The Zebracorns, NCSSM), Ryan Greenblatt (The Zebracorns, Brown University) | Bringing ROS to the Largest High School Robotics Competition | FIRST, or For The Inspiration and Recognition of Science and Technology, is an international organization focused on engaging students through STEM. This year, our team - The Zebracorns - was the first in the high school FIRST Robotics Competition (FRC) to control our robot entirely using ROS. In our presentation, we’ll introduce the unique challenges presented by FRC with restricted hardware options, time, and resources. We’ll talk about our motivation for implementing ROS, the specific application within FRC, and our ambitions for the future of ROS within the FRC community. | VideoSlides |
| 9:50 | Breandan Considine | Model-basedDuckietown: Software Infrastructure for Autonomous Robotics Development | Duckietown is an international research and education platform for autonomous vehicles, incorporating a physical and a virtual town populated with miniature autonomous cars. In this tutorial, we’ll take a grand tour of Duckietown and explore how machine learning, simulators and container technology are transforming robotics development. Attendees will learn how to train a self-driving vehicle using the Duckietown Gym environment and a little help from ROS. | VideoSlides |
| 10:10 | Risto Kojcev (Erle Robotics), Alejandro Hernández Cordero (Erle Robotics), Asier Bilbao Calvo (Erle Robotics), Irati Zamalloa Ugarte (Erle Robotics), Yue Leire Erro Nuin (Erle Robotics), Víctor Mayoral Vilches (Erle Robotics) | Accelerated robot training through simulation with ROS and Gazebo | Rather than programming, training allows robots to achieve behaviors that generalize better. However, such training requires big amount of experimentation which isn’t always feasible for physical robots. In this talk, we discuss gym-gazebo , a toolkit for developing and comparing reinforcement learning algorithms using ROS and Gazebo. In contrast to other alternatives, gym-gazebo applies roboticists’ tools, simplifying the development and deployment processes on real robots. We discuss the architecture of gym-gazebo and introduce early results that demonstrate that, for simple tasks, gym-gazebo accelerates the robot training time by more than 33% while maintaining same levels of accuracy and repeatability. | VideoSlides |
| 10:30 | Everyone | Coffee Break | ||
| 11:00 | Esteve Fernandez | ROS2 for Android, iOS and Universal Windows Platform: a demonstration of ROS2’s portability, and cross-platform and cross-language capabilities | ROS2’s layered architecture and adherence to modern C and C++ standards has enabled the support for the three major PC operating systems from the start. This talk will show that this design is flexible enough to have ROS2 running on architectures other than x86 (e.g. ARM) and on non-traditional platforms, such as Android, iOS and the Universal Windows Platform, while reusing a common codebase to all client libraries. Additionally, we’ll see how to write a generator and a client library through three examples (Java, Objective-C and C#) and how to integrate them into the ROS2 build system. | VideoSlides |
| 11:30 | Fabio Federici (United Technologies Research Center Italy), Giulio Mose’ Mancuso (United Technologies Research Center Italy) | Integrating ROS and ROS2 on mixed-critical robotic systems based on embedded heterogeneous platforms | This talk will focus on the analysis and tradeoff exploration of a hardware and software architecture for mixed-critical robotics applications. Specifically, we will consider the use of heterogeneous hardware platforms (i.e. System on Chip including CPUs, GPUs and FPGAs) , and we will define a software architecture allowing a seamless exploitation of these platforms under mixed real-time, safety and reliability requirements. We will show how ROS can be safely i ntegrated into this platform by adopting an appropriate resource partitioning. Possible benefits of using ROS2 in order to achieve predictability and real-time performance will also be analyzed. | VideoSlides |
| 11:40 | Iñigo Muguruza Goenaga (Erle Robotics), Juan Flores Muñoz (Erle Robotics), Víctor Mayoral Vilches (Erle Robotics), Loïc Dauphin (INRIA), Emmanuel Baccelli (INRIA), Cedric Adjih (INRIA) | Towards ROS 2 microcontroller meta cross-compilation | We introduce an extension of ament that proposes a new way of porting ROS 2 packages without modifications to microcontrollers. Our prototype cross-compiles packages into RTOS-specific projects that can later be compiled for the target platform. This work leads towards a ROS 2 meta cross-compilation setup that enables the use of microcontrollers as first-class participants of the ROS ecosystem while configuring their structure to meet the requirements by selecting through a variety of aspects such as RTOS, middleware, etc. We demonstrate this arrangement for two RTOSs, namely RIOT and NuttX, with microcontrollers as tiny as 32Kb RAM and 256Kb Flash. | Video |
| 11:50 | Minggang Wang (Intel Corporation), Kaining Yuan (Intel Corporation) | Node.js Cl ient & Web Bridge Ready for ROS 2.0 | The rclnodejs the Node.js* client for ROS 2.0, which is one of the first available clients . It enables developers to write robot applications in JavaScript* language, besides the hundreds of thousands of npm* packages can be leveraged. Also, with the new package ros2-web-bridge, browser based applications can inter-operate the ROS nodes. In this presentation, we will discuss the values of web technology for robotics application, then the design and architecture of rclnodejs and ros2-web-bridge will be introduced. Lastly, we will demonstrate how to write and deploy a web based ROS 2.0 application. | VideoSlides |
| 12:10 | Alejandro R. Mosteo (Centro Universitario de la Defensa, Zaragoza), Fabien Chouteau (AdaCore), Danilo Tardioli (Centro Universitario de la Defensa, Zaragoza), Yannick Moy (AdaCore) | RCLAda: the Ada client library for ROS2 | Ada is a general-purpose, strongly-typed, compiled language with a focus and long tradition in the embedded, high-integrity, mission-critical domains. This talk will present a newly available Ada binding to the ROS2 RCL library, rclada, which enables the writing of ROS2 nodes using the Ada language. Furthermore, the talk will describe, using ROS2 examples written with rclada, features of the language that make it an idoneous fit for robotics projects, and that may particularly interest experienced developers in other languages with similar target domains such as C++. | VideoSlides |
| 12:30 | Everyone | Lunch | Sponsored by Magazino | |
| 13:50 | Levi Armstrong (Southwest Research Institute), Jonathan Meyer (Southwest Research Institute) | Optimization Motion Planning with Tesseract and TrajOpt for Industrial Applications | Southwest Research Institute launched an internal R&D project to integrate the existing motion planner TrajOpt (Trajectory Optimization for Motion Planning) into ROS. TrajOpt was created at UC Berkeley as a software framework for generating robot trajectories by local optimization. The integration of TrajOpt necessitated new capabilities that spawned the creation of several new packages: tesseract and trajopt_ros . The tesseract package contains a new lightweight motion planning environment designed for industrial application, while the trajopt_ros package contains packages specific to TrajOpt. We will demonstrate how these tools complement existing planners, like OMPL or Descartes, to solve complex problems quickly and robustly. | Video |
| 14:20 | Michael ‘v4hn’ Görner (Hamburg University), Robert Haschke (Bielefeld University) | MoveIt! Task Planning | The MoveIt Task Constructor (MTC) is a recent extension of MoveIt to enable manipulation task planning. It provides a system to design, visualize and plan task-level robot trajectories composed of multiple interdependent sub-stages. In contrast to the previous pick and place manipulation module, it supports users in setting up custom task descriptions in few lines and provides debugging and visualization support along the way. | VideoSlides |
| 14:40 | Dave Coleman (PickNik Consulting) | Easy Robot Software and the MoveIt! Setup Assistant 2.0 | As the capabilities of robotic software frameworks increase, the setup difficulty and learning curve for new users also increase. If the entry barriers for configuring and using the software on robots is too high, even the most powerful of frameworks are useless. A growing need exists in ROS to aid users in getting started with, and customizing, the software framework as necessary for particular robotic applications. In this talk, we will explore the MoveIt! Motion Planning Framework’s approach to assisting new users with setup will be used as a case study into one of the challenges for pushing robotics forward. | VideoSlides |
| 15:00 | Chris Lewis (Southwest Research Institute), Jonathan Meyer (Southwest Research Institute) | Where’s My Camera | The industrial_calibration package suite in ROS-Industrial intends to provide mechanisms to automatically intrinsically and extrinsically calibrate cameras as part of a ROS installation on the factory floor. However , it has proven difficult to use. Several new nodes provide easy access to the most common calibrations including intrinsic, camera to robot tool, camera to robot and camera to camera. In addition, a new repository called robot_cal_tools provides a suite of off-line tools for creating new calibration routines, and for analyzing calibration results. This presentation describes how to use these new tools and accuracy you can expect. | VideoSlides |
| 15:20 | Everyone | Lightning Talks II | 1 - Introducing the Robotics Capture The Flag (video) (slides) 2 - IMcoders, IMU-based encoders (video) (slides) 3 - JdeRobot VisualStates: Visual tool for generating automata based robot behaviors (video) (slides) 4 - Mamihlapinatapei, on EusLisp (video) (slides) 5 - RaCER (video) (slides) 6 - ROSIN Funding for ROS developers and users (video) (slides) 7 - ROS and Construction Robots (video) (slides) 8 - ROS goes underground - Julius a mining robot (video) (slides) 9 - fmi_adapter - using Functional MockUp Units with ROS (video) (slides) 10 - The Autonomous Tree Pruning UAV (video) (slides) 11 - Open Source 3D Obstacle Recognition Applied to a Power Line Robot Inspection (video) (slides) 12 - 3DCoAutoSim (video) (slides) 13 - Low bandwidth driver for Universal Robots (video) (slides) 14 - Gauss6 A Six-axis ROS Arm (video) (slides) |
|
| 16:05 | Everyone | Coffee Break | ||
| 16:35 | Sarika Ramroop (University of the West Indies),Shivan Ramdhanie (Virtana TT Ltd.) | Integrating an Inertial Navigation System with the ROS Navigation Stack | This discusses the workflow used for integrating an Inertial Navigation System (INS) with the ROS Navigation stack. It involves taking GPS fix messages and Imu messages and converting from latitude, longitude, altitude (lla) format to the the map’s cartesian coordinates. This is then used to create the transforms between the global and local frames, and hence, the robot’s location within them. | VideoSlides |
| 16:45 | Andrew Blakey (Locus Robotics) | Unleashing the GIS Toolbox on Real-Time Robotics | Mobile robots generate large volumes of spatial data full of valuable information. In the ROS community we tend to reach for a handful of familiar toolsets and methodologies when asking spatial questions, leaving over 50 years of Geographical Information Systems (GIS) expertise untapped. This presentation introduces QGIS ROS, a QGIS plugin and custom data provider that enables first-class access of real-time and historical ROS data inside a complete GIS environment. | VideoSlides |
| 16:55 | Brian Cairl (Fetch Robotics Inc.) | Deterministic, asynchronous message driven task execution with ROS | This presentation discusses the building of a task-driven, message-synchronized execution abstraction layer on top of ROS’s message passing, aiming to mimic a deterministic system while continuing to use a non real time operating system. We discuss the overall structure of the framework, as well as the details of various capture and synchronization policies, and the profiling and diagnostics of task executions. We further discuss methods for using simulators and bagfiles in conjunction with the synchronization framework for repeatable testing and issue reproduction. | VideoSlides |
| 17:15 | Sebastian Pütz (Osnabrück University), Thomas Wiemann (Osnabrück University) | Introducing Tools for Storing, Rendering and Annotating Triangle Meshes in ROS and RViz | We introduce a set of tools to make 3D environment mesh representations accessible and manageable in ROS. We provide RViz tools to visualize and annotate huge meshes in combination with generated textures and different cost layers, which are based on the geometric analyses of the environment, or which represent different sensor readings, e.g. RGB image or even hyper-spectral image textures. | VideoSlides |
| 17:35 | Juraj Oršulić (Faculty of Electric Engineering and Computing, University of Zagreb) | Deterministic reversible debugging of ROS nodes with Mozilla rr | In this talk we give a short introduction to Mozilla rr, a powerful C/C++ debugging tool for Linux, and show how to more effectively debug ROS nodes by republishing messages from rr recordings. Mozilla rr serves as a gdb (GNU debugger) replacement which efficiently records the execution of a process and then provides repeatable deterministic debugging of the recording, enabling a very powerful debugging experience with reverse execution. | VideoSlides |
| 17:45 | Levon Avagyan (Fetch Robotics) | Hermetic Robot Deployment Using Multi-Stage Dockers | While encapsulating a ROS based stack in a docker has long been possible, the recent addition of multi-stage builds to Docker has made it feasible to build and deliver an entire robot deployment in a much more simple, concise, and maintainable way. This presentation will talk about methods to leverage multi-stage builds to improve build times, image sizes, and pipeline simplicity while maintaining consistency between development, testing, and production environments. | VideoSlides |
| 18:05 | Ryan Gariepy | Closing Remarks | VideoSlides | |
| 18:15 | Fin |
----