Writing robot control software is a challenging task. The mission of ros_control is to lower the entry barrier for exposing hardware to ROS, and to promote the reuse of control code in the same way that ROS has allowed for higher-level code.This talk presents an overview of the project, and the different tools it makes available to expose your robot to software, control it, and interface it with higher-level software. These tools include simulation backends, ready-to-use controllers, controller lifecycle management as well as building blocks for setting up new robots and writing custom controllers in a hardware-agnostic way.

The aim of this talk is to describe the implementation and deployment of dual arm assembly tasks using ROS. The development was done in the research project PRACE with an industrial mobile manipulation system consisting of the Fraunhofer rob@work 3 platform and an ABB dual arm concept robot. The experiences in planning and control for dual arm manipulation with ROS-Industrial drivers and MoveIt! are detailed and discussed. Furthermore, the integration of proprietary industrial libraries into a deployable ROS system is covered.

Recently three new physics engines (Bullet, Simbody, and DART) have been added to the Gazebo simulator to complement the default Open Dynamics Engine (ODE). Each physics engine offers a different set of trade-offs between accuracy, robustness, and speed. This presentation will compare the trade-offs of each physics engine in both simple physical tests and behavioral tests related to task level robotics simulation.

This talk will describe the details around ROS 2, provide rationale for selecting the underlying middleware and describe the development process and progress. You should gain a better understanding what to expect from ROS 2 and how you can contribute to the development to ensure that the next generation of the software meets your use cases.

For the European Robotics Challenges EuRoC several industrial robotic tasks are to be mastered. During qualifying stage, the challengers are asked to solve fully simulated challenges. The aim of this talk is to describe the development of a simulated human-robot collaboration task using ROS/gazebo. Furthermore several backend developments to allow controlling the scenario, logging and evaluating relevant data and automatic assessment are detailed and discussed.

A short review of the current status of ROS on ARM, what works, what is rumored to work, and what doesn’t work. Possibly also some photos and short descriptions of robots using ARM boards, if I can get them.

The rosserial package has humble origins as avr_bridge, a simple way to add DIO and low-cost I2C sensors to ROS. But rosserial today is far from a toy—it’s grown significantly beyond those beginnings, and we’d love to show some of what we’re doing with it.Clearpath Robotics now ships rosserial MCUs on Grizzly, Kingfisher, and a new, upcoming ground vehicle in the sub-Husky class. We’ve used rosserial on Windows and integrated it with ros_control. We think rosserial is great for giving the ROS user and developer maximal visibility into what is happening in all parts of the system.

This talk will give a high level overview of the various packages that exist for continuous integration, testing and build systems with the ROS ecosystem. It will discuss some of the difficulties these tools have, particularly in commercial environments and then focus on our recent efforts with the BuildBot-ROS system.

This talk will be a guided tour through the future ROS developer workflow, as we envision it. You will get a better picture how the different aspects of the ROS 2 development fit together and how they serve together to improve the developer experience. The presentation will get you excited to use these new tools.

This talk will present how ROS is being used for automation in the mining and heavy machinery industry. The i2Mine (Innovative Technologies and Concepts for the Intelligent Deep Mine of the Future) project, funded by the European Commission’s 7th Framework Programme, aims to bring to fruition the concept of an invisible, zero-impact mine which is safer and more productive. Varying levels of automation, from remote control to fully autonomous mining operations are being developed. ROS is used to fusion data from different commercial available sensors and to control a mobile test platform for the mine of the future.

In this talk, we will present our experience developing an advanced pick and place robot for surface mount transistor (SMT) printed circuit board (PCB) manufacturing. We will look at how we used ROS to automate the assembly setup-process and how we perform pick and place operations using a three axis gantry. In particular, we will relate to our experience using the TinyG, an open-hardware / open-software motion controller board in ROS. Additionally, we will discuss the design decisions that led us to choose ROS for industrial robotics control.

Using ROS in space systems has unique challenges including radiation, communication, and safety. Space systems must be somewhat error tolerant to radiation bit flips, must communicate over significant lag, restricted bandwidth and dropouts, and must operate in a known safe manner. This talk will provide an overview of Robonaut 2’s mobility upgrade. It will include an overview of our safety and communication systems and will discuss the challenges we faced and how we used ROS and OROCOS to solve them.

The ROS ecosystem is an interconnected web of packages, nodes and people with no efficient means to compare, assess or visualize them. We’ve developed a set of tools consisting of various metrics, a data visualization web app, and an active monitoring system. With these tools, we aim to elucidate the current state of the ecosystem as well as to determine where community efforts should be directed. We also want to encourage the community to provide input on potential applications, additional metrics, and further improvements to optimally address the needs of the ROS ecosystem.

The recent DARPA Robotics Challenge trials demonstrated many of the barriers of the current robotic technology, the complexity of the system development and the effort required to reach the performances needed to address practical tasks within unstructured environments. Both software and control architectures have major effects on the final robot capabilities. Designing effective control architecture for complex robotic systems, such as humanoid robots, still remains a challenge requiring to deal with the integration of a large number of software components. Considering the interest of the robotics community on humanoid robots, the aim of this paper is to discuss and highlight the pros and cons of the different control approaches in humanoid platforms. Following this, we introduce the control architecture developed, based ROS and YARP, for tele-operation of the humanoid robot COMAN.

Navigating is one of the key tasks for mobile robotics, and the Navigation Stack was one of the major milestones in the development of ROS. This talk will cover the central components of the system (costmaps and planners) and how they fit together. The talk will also include recent developments to extend ROS navigation to more use cases by making the components more extensible. Afterwards, users will have a better understanding of how to effectively configure their robot for their specific navigation needs.

The objective of the CompLACS research project is to decompose robotic problems into smaller, modular sub-problems that can be addressed using machine learning theory. This presentation will highlight the architectural challenges that emerge from this process, by focussing on work carried out on a swarm of quadrotors over the last year. In particular, we describe a ROS-based architecture called CRATES for streamlining the transition from simulation to experiments. We include results from software and hardware field trials in which path integral control theory is used to coordinate the behaviour of multiple quadrotors.

The new ROS-Industrial Calibration package provides a suite of calibration capabilities which encapsulate best practices and convenient interfaces to common robot and camera calibration tasks. It is scriptable and flexible to solve arbitrary combinations of multi-view intrinsic and extrinsic parameter optimization problems. It automates calibrations so that users may execute them without requiring significant domain knowledge. Two example calibrations will be described: 1) intrinsic and extrinsic calibration of a camera using a target positioned by a robot; 2) extrinsic calibration of a camera mounted on the end-effector. In both cases, MoveIt! provides collision free motion planning for the robot.

In this workshop, we will demonstrate how MATLAB® interacts with the Robot Operating System (ROS) using a new I/O Package. The package provides an API for creating ROS nodes in MATLAB that operate and communicate based on ROS’s publisher/subscriber mechanism. The ROS I/O Package has the following key features: (1) Enable creation of ROS nodes, publishers, and subscribers directly from MATLAB; (2) Enable creation of ROS messages from MATLAB; (3) Enable publishers to publish MATLAB data to their advertised topics; (4) Enable subscribers to execute arbitrary, user-defined MATLAB functions when messages are received; (5) Enable launching of a ROS Master in MATLAB.

Although we may like to think that the bulk of our development time is spent on things like developing new algorithms, much of that time actually goes to debugging, data introspection, and testing out small ideas. This can require lots of one-off scripts or time in an interactive shell – neither of which are well-supported by ROS. This talk will (re)introduce rosh, a python-based interactive shell and scripting environment for ROS. rosh basics will be discussed, including interaction with topics, services, geometry, visualization, and more, and how to use these to easily perform high-level robot control.