ros-industrial: leveraging open- source to force … presentations/ros-indust… · ros-i core...
TRANSCRIPT
ROS-INDUSTRIAL: LEVERAGING OPEN-SOURCE TO FORCE ADVANCED TECHNOLOGIES IN MANUFACTURING Matthew M. Robinson
SwRI: Deep Sea to Deep Space
SwRI Characteristics • Est. 1947 • San Antonio, Texas, USA • Independent, Not for profit • Applied RDT&E Services • Physical Science and Eng. • FY 2017 Revenue: $530M
Alvin submersible
New Horizons, Pluto
Manufacturing Robotics & Technologies Custom Robotics
Advanced Robotic Software Industrial Automation and Controls
Machine Vision and Perception
System Integration
ROS: Robot Operating System – Why?
• Open source (BSD)
• Established to keep robotics researchers from “re-inventing the wheel”
• Maintained by Open Robotics – 10 years strong!
• Reusable software components
• >1,000,000 users downloaded/mo. 1
1. http://download.ros.org/downloads/metrics/metrics-report-2017-07.pdf
Research Robotics • Reinvention of the Wheel • Little Commonality • Short Lifespan • Inability to Compare Results ROS Solves These
ROS: Robot Operating System Notable Users
2018: ROS-I
ROS-I Timeline
7
2010
SwRI Adopts ROS
2011
ROS-I Inception
2012
ROS-I Repo Launch
2013-2016
RIC Launch & Growth
Robotics Automated Coating Stripping Sys.
MR ROAM Mobile Robot
Robotic Workcell Visualization
What Can ROS-I Do?
Tech Vision Supported by Industry
ROSin – What is it? rosin-project.eu
Pitch to the funding agency (EC):
– “sweat equity” of OSS: those who put the work have a say
– instead of funding yet another framework, foster EU’s role in ROS with public €
4-years, ~8 million EUR H2020 project (1.2017-12.2020)
– Builds upon what exists; sustainable results after its completion
– Key actions to make ROS better, business friendlier, more accessible
– (Extra goal:) cluster other EU-based publicly funded activities using ROS
This project has been funded by
the European Union‘s Horizon2020 research and innovation programme under grant agreement No 732287
Key Actions to Make ROS:
Software Quality
ROS-I best practices and
tools: continuous integration,
unit testing, code reviews
ROSIN further improves
on them with code
scanning, automated test
generation, model-in-the-
loop testing
New components + path for exploitation
Education
Educate students:
summer schools
Train professionals:
ROS-I academy
Open Call to fund
your ROS education
initiative
3.5 Million € available to
third parties for
ROS-Industrial development
Develop missing components
or improve existing ones
Commercial release template
(licensing, etc)
better business friendlier more accessible
Evolution of Automation • Recent reports highlight evolution of automation
• Perception is that Manufacturing Work is nearer to complete automation
• For larger fabrication scenarios this work may be more appropriately defined as ‘Unpredictable Physical Work’
• This is where the development of ROS-Industrial plays a key role
Attacking End-User Opportunities Legacy Automation Support Costs
Moving of Variable Costs
Enabling Value Stream Optimization • The automated solution is only as successful as its inputs, supporting process steps, reliability, and
quality of output
• Merging ROS-Industrial and the initiatives around Digital Manufacturing/IoT, improved efficiencies in operations can be realized
Op 10 Op 20 Op 25 Op 30 Op 35 Every Step Every Time
Only Processing what is Needed
Completed Assembly
Robotic Blending – A Stepping Stone to Intelligent Agility
Opportunity 73% ROI based on overall part processing area efficiency improvements via reduction in variable labor
Scan-N-Plan Foundation
Target Adopter is a Low-Lot High Mix Manufacturing Site
Tech Demonstration of Robotic Blending Milestone 4 https://youtu.be/PWCpehyKnTY
• No tie-back to CAD – enable lower overhead for high-mix business
• Simple user interface
• No complex tooling, or need to precisely load parts
• Feature based processing
• Works on broad array of existing hardware – deployed on 10 year old existing robotic platform
• Increase in efficiency through redistribution of work across work area and a reduction in manual handling of parts
End User Value Case for Blending – Value Stream Approach
Realized a headcount reduction through work burden redistribution realizing 73% IRR and $1M NPV
A5 – Agility in Aerospace Applications
• Intelligent Framework, Adaptable to Many Processes
• Hardware Agnostic • Initial Pilot: Sanding for Paint
Prep, Composite Repair, NDI • High Value to Sustainment
Community
• The end-user
• Ability to Mock Up Capability with Modular Components and Incorporate IP
• Increased Agility and Value Proposition in Delivering IP more efficiently
• The deployment team (Internal or Integration Partner)
• Quicker Development of Capability
• Lower Cost in Getting to Something Viable
• Lower Cost to Support
How do we get to the value?
https://github.com/swri-robotics/euler
Bastian Solutions-ProMat April 2017
Courtesy of Bastian Solutions
Near-Term for Realizing Value Integrator or “Solution Provider” Deployment Model
Leverage the Integrator to Deliver Capability and Realize Value Sooner Fact: At many Industry Sites, Integrator Personnel our filling Operational Support Roles
Efficient Interoperability • 2017 Awarded NIST Grant building on 2013 work to
create “generic bridge” between MTConnect and ROS
• Break down the well-documented language barrier in factories between assets, facilitate many-to-many interoperability across the factory
• Enable for industry-wide adoption of open-source technologies/standards
– Providing a use-case(s) and testbed showcasing lower cost solutions for comprehensive factory floor integration
– Open paths for small- and medium-sized manufacturers
Photo courtesy of GE Power
Photo courtesy of motorman.com Y-blog
2018 IMTS Emerging Technology Center
Closed Loop Agile Process
• CAD+Scan to Path
• Closed Loop QA
• Improved Order to Delivery
• Improved Product to Market
• Reduced Losses during Product Change
• Enables ability to apply novel optimization and deep learning techniques for dynamic continuous improvement
PLM – Model with Embedded Features – Modeled Welds; Machine Features
MRP and Work Orders/Execution Systems
Source: creaform3d.com
Organization’s manufacturing IP/Process Data
Source: AdvancedManufacturing.org
Process Data to Drive Optimization
3D Printing/Additive Manufacturing • Leverage the Perception and Path Planning Capabilities and Tools within ROS-I to drive adaptable
3D Printing
Algorithms to adapt path plans to perceived environments/contours
New ROS Additive Manufacturing Package http://wiki.ros.org/ros_additive_manufacturing
https://www.youtube.com/watch?v=6U0zRbDbjQs
https://i.ytimg.com/vi/xauRUR6OK4M/maxresdefault.jpg
http://robots.iaac.net/images/7.jpg
Large Scale 3D Printing with Mobile Robots! Source: https://techcrunch.com/2017/08/29/3d-printed-space-habitats-earn-400k-in-prizes-at-nasa-competition/
ROS-I Core Technical Contributions to Open-Source Robotic Path Planning for Geometry-Constrained Process
• Framework (Plumbing)
– Plug-in Based
• Feature Extraction
– PCL Cluster Extraction
• Process Planning
– Raster path generation given a polygon mesh
• Sequence Planner
– Closest
• Trajectory Planning
– Descartes
Meshing, Process & Sequence Planning
Github
https://github.com/ros-industrial/noether
PCL Afront Mesher
https://github.com/PointCloudLibrary/pcl/pull/1996
ROS-I Core Technical Contributions to Open-Source
• Intelligent Part Reconstruction – TSDF + Next Best View (NBV) implementation for creation of meshes on featureless
or highly spectral parts
– Targeting path planning for high spectral/featureless parts
– Improved detail and efficiency in creation of meshes with greater level of detail
• Trajopt Integration – Cal Berkeley Trajectory Optimization for Motion Planning
– Recently released to open-source: https://github.com/ros-industrial-consortium/trajopt_ros
– Improved optimization-based path planning with multitude of cost functions
– Goal is to leverage for industrial manipulator applications
For the Developer – ROS Qt Creator
• Qt Creator Plug-in design specifically for ROS
• ROS Features
– Build System
• CatkinMake
• CatkinTools
– Run Configurations
– Automatic sourcing of workspace
– Integrated terminal
– Templates
• Developer
– Levi Armstrong
https://github.com/ros-industrial/ros_qtc_plugin
Profitability through Flexibility
• Demonstrated Cases of Impact on Improved Utilization of Assets
– New Capability in Legacy Assets
– Increased Flexibility
• Reduction in Non-Value/Low-Value Added Variable Labor
– Often high risk jobs
– Typically tasks subject to over-processing/little control
• Flexibility to be able to shift production closer to Market as demand shifts geographically; optimize outside the four walls
• Reduced overhead to support due to greater capability in variation management
• Improved accessibility to technology for the SME
The Newest MII Focusing on Robotics
Some objectives of the new institute: • Supporting advanced robotics capabilities for
manufacturing • Standardizing interfaces for cross-platform
compatibility • Modularizing and scaling components to larger
systems • Enabling a collaborative development environment • Developing the workforce through training
curriculum and hands-on classes • Transferring technology via open-source license • Providing affordability for small and medium
enterprises
SwRI and ROS-I Role: Ensuring synergy with global developments for efficient
leverage of resources; effective leverage of open source; sustainability
Resources
• ROS-Industrial • Home: rosindustrial.org • Documentation: wiki.ros.org/industrial • Code: https://github.com/ros-industrial;
https://github.com/ros-industrial-consortium • Training: http://ros-
industrial.github.io/industrial_training/ • ROSin: http://rosin-project.eu/
• Upcoming Events (https://rosindustrial.org/events-summary/)
