ROSCon 2016视频和幻灯片发布
By Tully Foote on 十月19，2016 7:28 AM
全部PPT下载地址：http://pan.baidu.com/s/1gf2sn2F
ROSCon在各方面都创下了历史新高，超过450名与会者，比去年的赞助增加了60％。
感谢大家来和你的支持！ 并感谢我们的赞助商的财务支持，使会议成长！
我们很高兴地宣布，我们已经发布了关于该计划的所有会谈的录音。 你可以找到它们链接到：
http://roscon.ros.org/2016/#program
我们也收集了大多数演示者的幻灯片，以及链接到那里。
如果你想单独浏览视频，你还可以在这里找到所有56视频：
https://vimeopro.com/osrfoundation/roscon-2016
具体链接如下：
由于技术问题，将无法使用直播。 在10月20日之前，会议网站上将提供演示文稿的存档视频。
视频记录和归档由支持。
下面是谷歌翻译，有译文和原文：
1日，10月8日
（跳转到第2天）
| 时间 | 主持人（S） | 标题 | 抽象 | 资源 |
|---|---|---|---|---|
| 8:00 | 大家 | 开放注册 | 请提前到达，以便有时间来收集你的徽章和会议袋的介绍再开始。我们预期将用于在第一天登记的队列。 | |
| 9:00 | 布赖恩Gerkey | 开场白 | PDF格式视频 | |
| 9:10 | 施特菲Paepcke，路易丝Poubel（OSRF） | 什么在凉亭新功能？升级你的模拟用户体验！ | 凉亭是在ROS社区最常用模拟器之一。它一直处于开发在过去几年中它的最新版本，凉亭7，自带了新老用户的一致好评无数新的工具和功能。近日，凉亭发展，强调以用户为中心的设计和改进的可用性。更新不仅包括改进的GUI工具，并为新人们的文档，而且还工具，简化工作流程有经验的用户。我们探索新的功能，包括：模型编辑器，编辑器大厦，施加外力工具，记录和回放，模型校准和捕捉工具，摄像头的角度控制，绘图，内省和调试工具，等等。 | PDF格式视频 |
| 9:50 | 迪安娜罩，威廉·伍德尔（OSRF） | ROS 2更新 | 本讲座将总结取得自2015年总结过去的ROSCon更新将包括在此期间发布的阿尔法，对支持的实现和路线图的进展情况。讲座还将包括同时演示和标杆使用ROS 2的新功能和我们的经验亮点演示。 | PDF格式视频 |
| 10:30 | 大家 | 咖啡时间（休闲时光 | ||
| 11:00 | 穆昆达Bharatheesha（代尔夫特理工大学），鲁本伯格（代尔夫特理工大学），马腾德弗里斯（开发者，代尔夫特机器人BV）的Gijs范德霍伦（代尔夫特理工大学），马亭维斯（代尔夫特理工大学） | 计划，MoveIt取胜！ - 兼谈亚马逊采摘挑战了解到2016年 | 本演讲将重点放在一些关键MoveIt的！1）作出适当MoveIt：我们（团队代尔夫特的运动规划团队），其次为亚马逊采摘挑战赛2016年特别是以下几点将突出实践！从一个大组选项API选择; 2）面临的困难，例如与轨迹和碰撞Octomaps和相应的解决办法检查I / O同步; 3）未解决的问题（主要是用机械手驱动），而周围的机器人的关节限制规划; 4）与MoveIt OMPL策划师配置一般建议！ | PDF格式视频 |
| 11:20 | 马修艾米（雷诺SAS），迈克尔·劳尔（LAAS-CNRS），让 - 查尔斯·法布尔（LAAS-CNRS） | 对ROS自适应容错：一个基于组件的方法 | 其业务生命周期中每个系统的发展。面对变化（新的威胁，失败，更新）时仍然可靠的系统称为弹性。我们提出了一个方法，安全性和适应性的容错利用基于组件的软件工程技术为解决弹性计算的一个关键方面，即在线适应的容错机制。我们将展示如何这种方法可以在ROS实施，解释一些实现细节和不同的实验结果来验证的解决方案。我们也将讨论我们如何能够利用检查点技术，使活性氧主人允许崩溃 | PDF格式视频 |
| 11:40 | Niharika阿罗拉（取机器人） | 机器人校准 | 校准是几乎所有机器人的基本前提。我们已经建立了一个快速，准确，robotagnostic校准系统，校准机器人的几何形状，除了典型的相机内部函数和/或extrinsics。该系统可与各种特征的检测器可以用于更新成本函数，并使用CERES优化的凸优化。然后，系统会创建一个包含校准参数的更新URDF。本讲座将涵盖robotagnostic robot_calibration包的详细信息，并说明它在fetch_calibration包，它可以在短短3分钟校准上取机器人几十个参数的使用。 | PDF格式视频 |
| 12:00 | 伊尼戈Muguruza Goenaga，亚历杭德罗·埃尔南德斯科尔德罗，维克托·马约拉尔比尔切斯（厄尔机器人） | 该robot_blockly包：积木编程ROS | robot_blockly是一个ROS包，允许用户创建基于ROS的算法和行为，其抽象用块的复杂性。包的目的是通过功能块来隐藏编程机器人的复杂性。作为一个经验法则，平均博士生需要3周的学习ROS。这使得ROS编程为绝大多数无法访问。所述robot_blockly包旨在简化使用ROS到推杆概念性块一起的点的处理。 | PDF格式视频 |
| 12:20 | 大家 | 午餐 | ||
| 13:50 | Yoonseok杓，柴田义，Ryuwoon荣，林Taehoon（ROBOTIS） | 介绍Turtlebot3 | 该Turtlebot3有三个概念：一个小的，可扩展的，开放的机器人平台。新Turtlebot的目标是降低该平台的大小和成本，而且不会降低性能，功能或质量。首先，我们将提供小尺寸Turtlebot3是没有足够的书包一个手提箱运输。开源嵌入式主板，计算机，传感器：其次，我们将为机械结构定制了一些备选方案提供了可扩展性。我们将提出Turtlebot3的12例。第三，我们将提供开源的H / W，F / W和S / W。只要你想，你可以改变你的Turtlebot3设计。 | PDF格式视频 |
| 14:10 | 阿米特·莫兰，马修·Curfman（英特尔 - 感知计算） | 英特尔介绍机器人RealSense所有功能于一身的感知设备 | 如果创建敏锐的机器人将是超级简单？即插即用。你会怎么做？英特尔RealSense™技术是周边机器的视觉营造创新的硬件和软件的结合。除其他事项外，硬件包括深度摄像头和软件包括一套CV算法，比如人跟踪，现场捕获，目标识别......在这次演讲中，我们将介绍一种创新我们创新实验室开发的所有功能于一身的机器人感知设备它可以提高任何机器人项目的开发，使机器人的感知，以几乎为零的安装。 | PDF格式视频 |
| 14:30 | 维克托·马约拉尔比尔切斯，亚历杭德罗·埃尔南德斯科尔德罗，伊尼戈Muguruza Goenaga，伊拉蒂Zamalloa加特，兰德Usategui乌加特（厄尔机器人） | 引入H-ROS，硬件机器人作业系统 | H-ROS代表硬件机器人作业系统，以创造轻松的可重用和r econfigurable机器人的硬件组件，简化机器人的创建，并允许来自不同制造商的部分使用ROS 2作为其核心基础进行互操作的，目的是DARPA资助的项目。H-ROS将帮助企业创造本身说话ROS机器人组件。随着H-ROS，建设一个机器人有关放置H-ROS兼容的硬件组件在一起。搬运机器人将不会仅限于具有较高的技术技能，少数而是将扩展到大多数与不同的H-ROS部分类型的一般理解。 | PDF格式视频 |
| 十四时50分 | 大家 | 闪电会谈中，我 | 视频 1234567891011121314 |
|
| 15:35 | 大家 | 羽化我的鸟 | ||
| 16:05 | 大家 | 茶歇（一羽讨论鸟类的继续） | ||
| 16:35 | 米尔科博尔迪尼翁（弗劳恩霍夫IPA），肖恩·爱德华兹（西南研究院），保罗·埃文斯（西南研究院），粘土Flannigan（西南研究院）的Gijs范德霍伦（代尔夫特理工大学），保罗Hvass（西南研究所）乌尔里希·赖泽（弗劳恩霍夫IPA），尼古拉斯·姚（ARTC - A * STAR），闵灵禅（ARTC - A * STAR） | ROS-工业变成四个全球扩张 | 四年后，这是第一次由肖恩·爱德华兹推出，ROS-工业现在是一个全球性的努力，扩大和改善ROS通过了生产环境和工业设备。作为一个举措，它是由原始设备制造商和系统集成商，他们的利益代表，并请求是三个ROS-工业财团非营利管理范围内收集的财政支持，应用研究机构。作为一个软件平台，它是通过每月例会一个自由和开放的社区聚集发展，并通过联合开发模式运行，很像ROS开发商社区。讲座将展示真实世界的例子和当前努力扩大主动，继续处理技术和非技术问题。 | PDF格式视频 |
| 16:55 | 列维·阿姆斯特朗（西南研究所） | ROS Qt Creator的项目经理插件 | 该ROS Qt Creator的插件是专门开发的ROS通过简化任务和创造ROS工具，一个集中的位置提高开发人员的工作效率。由于它是建立在Qt Creator的平台上，用户都可以访问所有喜欢它的现有功能：语法高亮显示，编辑（C ++，Python等），代码完成，版本控制（GIT，颠覆等）调试器（GDB，国开行，LLDB等），等等。讲座将包括：一个描述和动机; 当前和未来功能的概述; 和示例如何使用插件来管理ROS的工作空间。 | PDF格式视频 |
| 17:15 | 特里夫Thomessen，拉斯洛·纳吉（PPM AS） | 直觉ROS界面：4.0的FlexGUI - 引进和工业应用 | 4.0的FlexGUI基于流行的Web技术：HTML5，CSS和JavaScript。这样，就可以在PC机上运行4.0的FlexGUI，Android设备，iPhone，Windows手机和一般用现代的浏览器在每个设备上，你会等皆正是他们每个人相同的用户体验。4.0的FlexGUI相通使用ROS，我们选择的中间件的工业4.0，物联网。加入会话，让我们自我介绍，请参见我们的一些工业应用，请问我们怎样可以帮助你和你的！ | PDF格式视频 |
| 17:35 | 保罗Hvass，肖恩·爱德华兹（西南研究所） | ROS-'X“ - 重点的倡议 | 你听说过“ROS-I”或“ROS-M”的项目？什么人？在各种商业市场ROS收益验收，ROS-'X“项目似乎来组织他们周围。这些市场包括工业，军工，农业，航空，汽车等。这种努力的概述，有些比其他人更成熟，将提交。这些项目形成各种理由进行讨论以及对更广泛的活性氧社会带来的影响。有于一体的专业ROS-'X的项目潜力巨大，但这些项目也有很大的责任，以确保更广泛的ROS项目继续，并很好的支持。 | PDF格式视频 |
| 18:00 | 大家 | 前台 | ||
| 20:00 | 鳍 |
2日，10月9日
（跳转至第1天）
| 时间 | 主持人（S） | 标题 | 抽象 | 资源 |
|---|---|---|---|---|
| 8:30 | 大家 | 开放注册 | ||
| 9:00 | 彼得FANKHAUSER（ETH苏黎世），基督教Gehring集团（苏黎世），圣雷莫Diethelm（ETH苏黎世），塞缪尔·巴赫曼（ETH苏黎世），达里奥Bellicoso（ETH苏黎世），马丁Wermelinger（ETH苏黎世），瓦西利奥斯Tsounis（ETH苏黎世），迈克尔· Bloesch（ETH苏黎世），菲利普Leemann（ETH苏黎世），百利Hottiger（ETH苏黎世），安德烈亚斯·劳伯（ETH苏黎世），多米尼克士（苏黎世），拉尔夫Kaestner（IBM研究苏黎世），马克Hoepflinger（armasuisse），马可·胡特（苏黎世联邦理工学院） | ANYmal在ARGOS挑战：从石油和天然气站点的自主检查工具和经验，腿式机器人 | 在ARGOS挑战，我们参与我们的石油和天然气站点的自主检查足机器人ANYmal。在本文中，我们在我们的工作中使用ROS分享我们的经验和介绍我们的工具，其中许多是开源的。ANYmal是电驱动的四足机器人，能够动态运行，攀越障碍，缩放楼梯。它采用板载激光测距传感器来精确定位和地图及其周边，能够在困难的环境中自主导航。对于ARGOS，ANYmal进行检查几个传感器，并使用计算机视觉监控现场的状态。 | PDF格式视频 |
| 9:40 | 德克·托马斯（OSRF） | 该ROS生成的农场 - 它可以为我做 | 这次报告的重点是Python包ros_buildfarm。提供的各种功能的详细这将提高理解ROS包从开发和持续集成在发布过程中的文档的开发过程中所描述。它会帮助你利用build.ros.org所提供的基础设施的ROS软件包的发展，提高您的工作流程。此外，它正在演示软件如何在本地应用或如通过特拉维斯运行自己的“工作”。 | PDF格式视频 |
| 10:20 | 迈克·珀维斯（机器人的ClearPath） | 与ROS捆绑可靠的部署 | 后在2015年的ClearPath正面临着部署危机，在各种状态下的软件去客户现场，非现场演示，内部演示和测试，和开发者工作站。有在现有工具的文化问题和技术的限制积聚。所有这些问题的解决一直在打造我们的机器人软件的全部在一个较大的工作空间（使用catkin_tools），然后船整个事情为一体的“肥肉”deb包。经验教训将提出，以建设为Ubuntu定制desktop_full束的简单例子一起 - 这应该是足够的kickstart任何人想建立一个类似的建设谁的努力。该演示将特别突出我们的贡献上游已在此开发工作的过程中取得了ROS的工具，在未来几个月内有望合并 | PDF格式视频 |
| 10:40 | 大家 | 咖啡时间（休闲时光 | ||
| 11:10 | 伊恩·陈（OSRF） | 多云，有仿真的机会 | DARPA的虚拟机器人挑战赛催生CloudSim的第一个版本。它提供了便利来自世界各地的参会团队在模拟同时竞争一个集中的平台。此后，各种项目纷纷要求对机器人竞赛类似的云托管环境的需要。在这份报告中，我们揭示了一个新的，重新设计的CloudSim的目标是成为在云中运行的机器人软件和模拟的通用工具。我们提出它的一些关键特性包括仿真实例共享，可重复使用的前端Web组件，一个WebGL的可视化客户端，安全和认证。 | PDF格式视频 |
| 11:30 | 里卡多·特列斯（该结构） | 机器人与标杆ROS | 机器人基准是一个尚未解决的问题。具有机器人标杆系统将加快机器人技术和研究开发，因为结果可能会针对相同的条件下相互比较，新的机器人作品可能最终打造建立在以前的结果。本次会议的目的是描述一个框架，一个普遍的基准系统基于模拟凉亭，在那里ROS有可能成为硬件和软件之间的通信标准中的核心作用的机器人。 | 视频 |
| 11:50 | 亚历克斯·亨宁（取机器人） | 物理持续集成 - CI实物机器人！ | 取机器人已经制定出一套完整的系统测试新的代码，包括多个真实的机器人持续集成（CI）系统。多ceilingmounted相机和onrobot rosbag录音允许开发人员在挖掘和发现，当诊断错误。本讲座将涵盖我们的机器人舰队的工具，我们已经用于部署方法和监控测试。它还将支付必要的运行真实世界测试的其他注意事项，包括使用物理持续集成，以验证bug修复的ROS导航堆栈的案例研究。 | PDF格式视频 |
| 12:10 | 大家 | 午餐 | ||
| 13:40 | 鲁芬怀特（美国加州大学圣地亚哥分校），摩根·奎格利（OSRF） | {，S}的ROS：保护ROS通过线路，在图中，并通过在内核 | 本讲座将展示SROS，提议除ROS API和生态系统的使用，以支持现代密码学和安全性的措施。虽然详细说明目前的进展，并进一步建议，每一个重大进步的一个工作演示将显示，包括：对所有消息，服务和参数运输overthewire密码，名称空间的访问控制实施图形政策/限制，最后过程中使用Linux安全模块的配置文件硬化节点的文件，设备，信号和网络上的主机上访问。您将获得在ROS和持续发展的脆弱性更清晰的认识，以确保机器人子系统的未来。 | PDF格式视频 |
| 14:00 | 亚当·东布罗夫斯基，拉法尔Kozik，Mateusz马西亚斯（PIAP） | 评估ROS2通信层的复原力 | 一对ROS2发展的主要原因是通信层DDS的迁移。机器人世界各地的用户，通过ROS缺点迄今气馁，最终可能考虑加入社区。对于所有那些谁看到的承诺，就有ROS2是否会兑现它的问题。我们引入ROS2通信可测量，可重复评价一个项目，在不到完美的网络。我们将目前从评估工具的最新成果，并得出结论：取得了哪些成果？有什么解决的最重要的差距？ | PDF格式视频 |
| 14:20 | 扬卡斯特森，阿克塞尔Rauschenberger（莱布尼兹汉诺威Universität大学） | RTROS - 一种实时扩展到机器人作业系统 | RTROS使ROS用户能够轻松端口或养成使用许多来自ROS生态系统带来的好处，同时满足硬实时限制ROS节点。测试时间Xenomai和抢占-RT库提供了确定性的沟通和核心功能开发的实时机器人系统。RTROS被设计为使得独立RTROS或ROS / RTROS系统是可能的。因此，对于非关键节点使用标准的ROS节点仍然是可能的。 | PDF格式视频 |
| 14:40 | 大家 | 闪电II会谈 | 视频 123456789101112131415 |
|
| 15:25 | 大家 | 羽化二鸟 | ||
| 15:55 | 大家 | 茶歇（一羽讨论鸟类的继续） | ||
| 16:25 | 德扬Pangercic，大卫球，马库斯Hoeferlin，丹尼尔·迪马可，斯特芬福克斯，托比亚斯Mugele，托比亚斯北斗七星，加布里埃尔Gaessler，安德烈亚斯迈克尔斯，桑迪普Bairampalli，莫里斯Gohlke，罗兰 - 科斯塔Tschakarow，阿莫斯·艾伯特（Deepfield机器人） | 农业机器人与ROS博世：从领域的互联网设备的互联网 | 博世启动Deepfield机器人正在为自动化的工厂表型的创新体系。我们的系统包括用于数据采集和检测和植物的分析处理流水线半自主机器人。我们的系统是在广泛ROS基础。本讲座将为我们专注于我们的表型的植物计数管道，机器人的软件和实时设计，对ROS bagfile管理，生命周期管理我们的最佳实践完整的系统的详细介绍，建农场，配置，ROS的框架内。本讲座将有兴趣使用ROS其他初创公司，并希望提高其质量水平等群体。到目前为止，我们已经处理过的数据的TB级27和多日检测超过200万株。 | PDF格式视频 |
| 16:45 | 德扬Pangercic（Deepfield机器人），丹尼尔·马可迪（Deepfield机器人），弗洛里安Friesdorf（Ternaris UG），马尔科Durkovic（Ternaris UG） | Bagbunker - 工具大容量数据存储，分析，查看和测试 | 管理大数据集是用于机器人专家一个具有挑战性的，但往往低估的任务，尤其是在户外应用如农业，水下和空间机器人，以及自动驾驶。我们已经发布了bagbunker，对于大型数据集的存储，管理和检索一个基于Web的工具，如rosbags。Bagbunker就像是一个照片应用程序，只需用一个不同的数据。它是专门设计来提高管理大型数据集的任务，也为数据库系统没有训练有素的工程师。以这种方式，它极大地通过保持中央控制下的数据并提供一个基于网络的接口进行过滤，检查，可视化和与数据交互加快开发和调试。在这种HANDSON会话参与者将学习如何有效地管理大型数据集，包括如何搜索和检索rosbags以及如何编写工作分析rosbags。此外，我们还将介绍如何使用bagbunker一个图像分类的回归测试。 | 视频 |
| 17:05 | 诺曼李，杨保罗（DJI） | 活性氧开发商提供强大飞行平台 | 演示文稿与其相应的ROS软件包航空机器人开发人员介绍了一种准确，稳定，可靠的飞行平台DJI矩阵的计算100在一起。开发人员可以监视和控制飞机的飞行行为在船上SDK包ROS的帮助，同时利用内置的智能导航模式打造自主飞行路径和演习。它的ROS包是功能强大且易于使用，它已经帮助许多矩阵的计算100开发商赚取他们的研究和开发过程中伟大的成就。 | PDF格式视频 |
| 17:25 | 瑞安Gariepy | 结束语 | PDF格式视频 | |
| 17:35 |
Locations
Registration: Grand Ballroom 3, Level 3
General Sessions: Grand Ballrooms 1-2, Level 3
Coffee Breaks: Foyer and Grand Ballroom 3, Level 3
Reception: Foyer and Grand Ballroom 3, Level 3
Lunch: Foyer and Park Ballrooms 1-3, Level 5Note: the exact scheduling of presentations is subject to change.
Day 1, October 8th
(jump to Day 2)
| Time | Presenter(s) | Title | Abstract | Resources |
|---|---|---|---|---|
| 8:00 | 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 | Opening Remarks | pdf video | |
| 9:10 | Steffi Paepcke, Louise Poubel (OSRF) | What’s new in Gazebo? Upgrading your simulation user experience! | Gazebo is one of the most used simulators in the ROS community. It has been under heavy development for the past few years and its most recent version, Gazebo 7, comes with myriad new tools and features for new and experienced users alike. Recently, Gazebo development has emphasized user-centered design and improved usability. Updates include not only improved GUI tools and documentation for new folks, but also tools that streamline the workflow for experienced users. We explore new features including: Model Editor, Building Editor, apply force tool, logging and playback, model alignment and snap tools, camera angle controls, plotting, introspection and debugging aids, and more. | pdf video |
| 9:50 | Deanna Hood, William Woodall (OSRF) | ROS 2 Update | This talk will summarize the progress made since the last ROSCon update in 2015. Summary will include the alphas released during that time, changes to supported implementations, and the roadmap. The talk will also include demonstrations of new features and highlights of our experiences while using ROS 2 in demos and benchmarking. | pdf video |
| 10:30 | Everyone | Coffee Break | ||
| 11:00 | Mukunda Bharatheesha (Delft University of Technology), Ruben Burger (Delft University of Technology), Maarten De Vries (Developer, Delft Robotics BV), Gijs van der Hoorn (Delft University of Technology), Martijn Wisse (Delft University of Technology) | Plan to Win with MoveIt! - Lessons learnt from the Amazon Picking Challenge 2016 | This presentation will focus on some of the key MoveIt! practices that we (motion planning team of Team Delft) followed for the Amazon Picking Challenge 2016. Particularly, the following points will be highlighted: 1) making appropriate MoveIt! API choices from a large set of options; 2) difficulties faced such as I/O synchronization with trajectories and collision checking with Octomaps and the corresponding solutions; 3) unsolved problems (mostly with robot driver) while planning around the joint limits of the robot; and 4) general recommendations for OMPL planner configurations with MoveIt!. | pdf video |
| 11:20 | Matthieu Amy (Renault SAS), Michaël Lauer (LAAS-CNRS), Jean-Charles Fabre (LAAS-CNRS) | Adaptive Fault Tolerance on ROS: A Component-Based Approach | Every system evolves during their operational lifetime. A system that remains dependable when facing changes (new threats, failures, updates) is called resilient. We propose an approach to safety and adaptive fault tolerance taking advantage of Component-Based Software Engineering technologies for tackling a crucial aspect of resilient computing, namely the on-line adaptation of fault tolerance mechanisms. We will show how this approach can be implemented on ROS and explain some implementation details and the result of different experiments to validate the solution. We will also discuss the how we can use checkpointing technologies to make the ROS master crash-tolerant | pdf video |
| 11:40 | Niharika Arora (Fetch Robotics) | Robot calibration | Calibration is an essential prerequisite for nearly any robot. We have created a fast, accurate and robotagnostic calibration system, which calibrates robot geometry in addition to the typical camera intrinsics and/or extrinsics. The system can be used with a variety of feature detectors to update the cost function and uses the CERES optimizer for the convex optimization. The system then creates an updated URDF containing the calibrated parameters. This talk will cover the details of the robotagnostic robot_calibration package and describe its use in the fetch_calibration package which can calibrate dozens of parameters on a Fetch robot in as little as 3 minutes. | pdf video |
| 12:00 | Iñigo Muguruza Goenaga, Alejandro Hernández Cordero, Víctor Mayoral Vilches (Erle Robotics) | The robot_blockly package: programming ROS with blocks | robot_blockly is a ROS package that allows users to create ROS-based algorithms and behaviors, abstracting its complexity using blocks. The aim of the package is to hide the complexity of programming robots via functional blocks. As a rule of thumb, an average PhD student takes 3 weeks to learn ROS. This makes ROS programming not accessible for the great majority. The robot_blockly package aims to simplify the process of using ROS to the point of putting conceptual blocks together. | pdf video |
| 12:20 | Everyone | Lunch | ||
| 13:50 | Yoonseok Pyo, Yoshihiro Shibata, Ryuwoon Jung, Taehoon Lim (ROBOTIS) | Introducing the Turtlebot3 | The Turtlebot3 has three concepts: a small, extensible, open robot platform. The goal of new Turtlebot is to reduce the size and cost of the platform without reducing capability, functionality, or quality. Firstly, we will provide small size Turtlebot3 that is enough to be transported without a carrying case in a school bag. Secondly, we will offer extensibility for customizing on the mechanical structure with some alternative options: opensource embedded board, the computer, and the sensors. We will present 12 examples of Turtlebot3. Thirdly, we will provide opensource H/W, F/W, and S/W. You can change your Turtlebot3 design as you want. | pdf video |
| 14:10 | Amit Moran, Matthew Curfman (Intel - Perceptual Computing) | Introducing Intel RealSense Robotics All-in-one Perception Device | What if creating perceptive robots would have been super easy? Just plug and play. What would you do? Intel® RealSense™ Technology is a combination of hardware and software for creating innovation around machine perception. Among other things, the hardware includes depth camera and the software includes a set of CV algorithms such as person tracking, scene capturing, object recognition… In this talk we will introduce an innovative All-in-one Robotics Perception Device developed in our innovation lab which can boost the development of any robotic project and enable robots’ perception, with virtually zero installation. | pdf video |
| 14:30 | Víctor Mayoral Vilches, Alejandro Hernández Cordero, Iñigo Muguruza Goenaga, Irati Zamalloa Ugarte, Lander Usategui Ugarte (Erle Robotics) | Introducing H-ROS, the Hardware Robot Operating System | H-ROS stands for Hardware Robot Operating System, a DARPA-funded project with the aim of creating easily reusable and r econfigurable robot hardware components to simplify the creation o f robots and allow parts from different manufacturers to interoperate using ROS 2 as its core basis. H-ROS will help companies create robot components that speak ROS natively. With H-ROS, building a robot is about placing H-ROS-compatible hardware components together. Handling robots won’t be restricted to a few with high technical skills but will instead be extended to a great majority with a general understanding of the different H-ROS part types. | pdf video |
| 14:50 | Everyone | Lightning Talks I | Videos 1 2 3 4 5 6 7 8 9 10 11 12 13 14 |
|
| 15:35 | Everyone | Birds of a Feather I | ||
| 16:05 | Everyone | Coffee Break (and continuation of Birds of a Feather discussions) | ||
| 16:35 | Mirko Bordignon (Fraunhofer IPA), Shaun Edwards (Southwest Research Institute), Paul Evans (Southwest Research Institute), Clay Flannigan (Southwest Research Institute), Gijs van der Hoorn (Delft University of Technology), Paul Hvass (Southwest Research Institute), Ulrich Reiser (Fraunhofer IPA), Nicolas Yeo (ARTC – A*STAR), Min Ling Chan (ARTC – A*STAR) | ROS-Industrial turns four and expands worldwide | Four years after it was first launched by Shaun Edwards, ROS-Industrial is now a worldwide effort to expand and improve ROS adoption in manufacturing environments and industrial equipment. As an initiative, it is supported financially by OEMs and system integrators, whose interests are represented and requests are collected within three ROS-Industrial Consortia managed by non-profit, applied research institutions. As a software platform, it is developed by a free and open community gathering in monthly meetings and operating through a federated development model, much like the ROS developers community. The talk will showcase real-world examples and current efforts to expand the initiative and to continue addressing technical and non-technical concerns. | pdf video |
| 16:55 | Levi Armstrong (Southwest Research Institute) | ROS Qt Creator Project Manager Plug-in | The ROS Qt Creator Plug-in is developed specifically for ROS to increase developers’ efficiency by simplifying tasks and creating a centralized location for ROS tools. Since it is built on top of the Qt Creator platform, users have access to all of its existing features like: syntax highlighting, editors (C++, Python, etc.), code completion, version control (Git, Subversion, etc.), debuggers (GDB, CDB, LLDB, etc.), and much more. The talk will cover: a description and motivation; overview of current and future features; and example of how to use the plug-in to manage a ROS Workspace. | pdf video |
| 17:15 | Trygve Thomessen, Laszlo Nagy (PPM AS) | The Intuitive ROS UI: FlexGui 4.0 – introduction and industrial applications | FlexGui 4.0 is based upon popular web technologies: HTML5, CSS and JavaScript. This way it is possible to run FlexGUI 4.0 on PC, Android, iPhone, Windows Phone and generally on every device with a modern browser, you will h ave exactly the same user experience on each of them. FlexGui 4.0 communicates using ROS, our choice of middleware for Industry 4.0, IoT. Join the session and let us introduce ourselves, see some of our industrial applications, ask us how could we help you with yours! | pdf video |
| 17:35 | Paul Hvass, Shaun Edwards (Southwest Research Institute) | ROS-‘X’ – Focused Initiatives | Have you heard of the “ROS-I” or “ROS-M” projects? What about others? As ROS gains acceptance in various commercial markets, ROS-‘X’ projects seem to organize around them. Such markets include industrial, military, agriculture, aerial, and automotive. An overview of such efforts, some more mature than others, will be presented. The various reasons these projects form will be discussed as well as the impact on the broader ROS community. There is great potential in specialized ROS-‘X’ projects, but these projects also have great responsibility to ensure that the broader ROS project continues and is well supported. | pdf video |
| 18:00 | Everyone | Reception | ||
| 20:00 | Fin |
Day 2, October 9th
(jump to Day 1)
| Time | Presenter(s) | Title | Abstract | Resources |
|---|---|---|---|---|
| 8:30 | Everyone | Registration open | ||
| 9:00 | Péter Fankhauser (ETH Zurich), Christian Gehring (ETH Zurich), Remo Diethelm (ETH Zurich), Samuel Bachmann (ETH Zurich), Dario Bellicoso (ETH Zurich), Martin Wermelinger (ETH Zurich), Vasilios Tsounis (ETH Zurich), Michael Bloesch (ETH Zurich), Philipp Leemann (ETH Zurich), Gabriel Hottiger (ETH Zurich), Andreas Lauber (ETH Zurich), Dominik Jud (ETH Zurich), Ralf Kaestner (IBM Research Zurich), Mark Hoepflinger (armasuisse), Marco Hutter (ETH Zurich) | ANYmal at the ARGOS Challenge: Tools and Experiences from the Autonomous Inspection of Oil & Gas Sites with a Legged Robot | At the ARGOS Challenge, we participate with our legged robot ANYmal in the autonomous inspection of oil & gas sites. In this presentation, we share our experience with using ROS in our work and introduce our tools, many of which are available open-source. ANYmal is an electrically driven quadrupedal robot, capable of dynamic running, climbing over obstacles, and scaling stairs. It uses onboard laser range sensors to precisely localize and map its surrounding, enabling the autonomous navigation in difficult environments. For ARGOS, ANYmal carries several inspection sensors and uses computer vision to monitor the state of the site. | pdf video |
| 9:40 | Dirk Thomas (OSRF) | The ROS build farm - what it can do for me | The presentation is focused on the Python package ros_buildfarm. The various features provided are being described in detail which will improve the understand of the development process of ROS packages ranging from development and continuous integration over the release process to documentation. It will help you to leverage the provided infrastructure of build.ros.org for the development of your ROS packages and improve your workflow. Additionally it is being demonstrated how the software can be applied locally or e.g. via Travis to run your own “jobs”. | pdf video |
| 10:20 | Mike Purvis (Clearpath Robotics) | Robust Deployment with ROS Bundles | Late in 2015, Clearpath was facing a deployment crisis, with software in various states going to customer sites, off-site demos, in-house demos & testing, and developer workstations. There was a build-up of cultural issues and technical limitations in the existing tooling. The solution to all of these issues has been to build the entirety of our robot’s software in a single large workspace (using catkin_tools), and then ship the whole thing as one “fat” deb package. Lessons learned will be presented, along with a brief example of building a customized desktop_full bundle for Ubuntu - this should be sufficient to kickstart the efforts of anyone else who’d like to set up a similar build. The demonstration will highlight in particular our contributions to upstream ROS tooling which have been made in the course of this development work, hopefully merged in coming months | pdf video |
| 10:40 | Everyone | Coffee Break | ||
| 11:10 | Ian Chen (OSRF) | Cloudy with a Chance of Simulation | The DARPA Virtual Robotics Challenge gave birth to the first version of CloudSim. It provided a centralized platform that facilitated teams of participants from around the world to compete in simulation simultaneously. Since then, various projects have demanded the need for a similar cloud hosted environment for robotics competitions. In this presentation, we reveal a new, redesigned CloudSim that aims to be a generic tool for running robotics software and simulations on the cloud. We present some of its key features including simulation instance sharing, reusable front-end web components, a WebGL visualization client, and security and authentication. | pdf video |
| 11:30 | Ricardo Tellez (The Construct) | Robotics Benchmarking with ROS | Benchmarking in robotics is an unsolved issue. Having a benchmarking system for robotics would speed up development of robotics and research since results may be compared against each other in same conditions, and new robotics works may finally build build on previous results. The purpose of this session is to describe a framework for a universal benchmarking system for robotics based on Gazebo simulations, where ROS has the central role of becoming the standard of communication between hardware and software. | video |
| 11:50 | Alex Henning (Fetch Robotics) | Physical Continuous Integration — CI on Real Robots! | Fetch Robotics has developed a continuous integration (CI) system for testing new code in a complete system including multiple real robots. Multiple ceilingmounted cameras and onrobot rosbag recordings allow developers to dig in and diagnose bugs when found. This talk will cover the tools and methods we have used to deploy and monitor tests on our fleet of robots. It will also cover additional considerations necessary for running real world tests and include a case study of using physical continuous integration to verify bug fixes to the ROS navigation stack. | pdf video |
| 12:10 | Everyone | Lunch | ||
| 13:40 | Ruffin White (University of California, San Diego), Morgan Quigley (OSRF) | {,S}ROS: Securing ROS over the wire, in the graph, and through the kernel | This talk will demonstrate the uses of SROS, a proposed addition to the ROS API and ecosystem to support modern cryptography and security measures. While detailing current progress and further proposals, a working demonstration of each major advancement will be shown, including: overthewire cryptography for all messages, services and parameter transport, namespaced access control enforcing graph policies/restrictions, and finally process profiles using Linux Security Modules to harden a node’s file, device, signal, and networking access on host machines. You’ll gain a greater awareness of the vulnerabilities in ROS and ongoing developments to secure robotics subsystems for the future. | pdf video |
| 14:00 | Adam Dąbrowski, Rafał Kozik, Mateusz Macias (PIAP) | Evaluating the resilience of ROS2 communication layer | One of the major reasons for development of ROS2 is the migration of communication layer to DDS. Robot users around the world, hitherto discouraged by the ROS shortcomings, might finally be considering joining the community. For all those who see the promise, there is a question of whether ROS2 will deliver on it. We introduce a project for measurable, repeatable evaluation of ROS2 communication in less than perfect networks. We will present the most recent results from the evaluation tool and draw conclusions: what has been achieved? What are the most important gaps to address? | pdf video |
| 14:20 | Jan Carstensen, Axel Rauschenberger (Leibniz Universität Hannover) | RTROS – A real-time extension to the Robot Operating System | RTROS enables ROS users to easily port or develop ROS nodes using many of the benefits from the ROS-ecosystem, while satisfying hard real-time constraints. Tested for Xenomai and preempt-RT the library offers deterministic communication and core functionality to develop a real-time robotic system. RTROS is designed such that standalone RTROS or ROS/RTROS system are possible. Thus, the use of standard ROS nodes for uncritical nodes is still possible. | pdf video |
| 14:40 | Everyone | Lightning Talks II | Videos 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 |
|
| 15:25 | Everyone | Birds of a Feather II | ||
| 15:55 | Everyone | Coffee Break (and continuation of Birds of a Feather discussions) | ||
| 16:25 | Dejan Pangercic, David Ball, Markus Hoeferlin, Daniel Di Marco, Steffen Fuchs, Tobias Mugele, Tobias Dipper, Gabriel Gaessler, Andreas Michaels, Sandeep Bairampalli, Maurice Gohlke, Roland-Kosta Tschakarow, Amos Albert(Deepfield Robotics) | Agricultural Robotics with ROS at Bosch: From the internet of fields to the internet of plants | Bosch Startup Deepfield Robotics is creating innovative systems for automated plant phenotyping. Our system comprises semi-autonomous robots for data acquisition and a processing pipeline for detection and analysis of plants. Our system is based extensively on ROS. This talk will provide a detailed description of our complete system focusing on our phenotyping plant counting pipeline, the robot’s software and real-time design, our best practices for ROS bagfile management, life cycle management, build farm, provisioning, within the ROS framework. This talk will be of interest to other startups using ROS and other groups wishing to increase their level of quality. So far we have processed 27 terabytes of data and detected over 2 million plants across multiple days. | pdf video |
| 16:45 | Dejan Pangercic (Deepfield Robotics), Daniel Di Marco (Deepfield Robotics), Florian Friesdorf (Ternaris UG), Marko Durkovic (Ternaris UG) | Bagbunker - Tool for Large Data Storage, Analysis, Viewing and Testing | Managing large datasets is a challenging, but often underappreciated task for roboticists, especially in outdoor applications such as agriculture, underwater and space robotics, as well as autonomous driving. We have released bagbunker, a webbased tool for storage, management and retrieval of large datasets such as rosbags. Bagbunker is like a photo app, just with a different data. It is specifically designed to improve the task of managing large datasets, and also for engineers not trained in database systems. In this way, it greatly speeds up development and debugging by keeping the data under central control and providing a webbased interface to filter, inspect, visualize and interact with the data. In this handson session participants will learn how to efficiently manage their large datasets, including how to search and retrieve rosbags and how to write jobs to analyse rosbags. Additionally we will also present how to use bagbunker for an image classification regression test. | video |
| 17:05 | Norman LI, Paul Yang (DJI) | A robust flying platform for ROS developers | The presentation introduces an accurate, stable and reliable flying platform DJI Matrice 100 together with its corresponding ROS packages to aero robotics developers. Developers can monitor and control aircraft flight behavior with the help of On board SDK ROS packages, while utilizing the built-in Intelligent Navigation Modes to create autonomous flight paths and maneuvers. Its ROS packages are powerful and easy to use, which have helped many Matrice 100 developers earn great achievement in their research and development process. | pdf video |
| 17:25 | Ryan Gariepy | Closing Remarks | pdf video | |
| 17:35 | Fin |
