FAIM2017 27-30 June 2017 – University of Modena and RE
Towards shared autonomy for robotic tasks in manufacturing
Presenting author: Sharath Chandra Akkaladevi, Profactor GmbH
Andreas Pichler, Markus Ikeda, Michael Hofmann, Matthias Plasch, Christian Wögerer, Gerald Fritz
FAIM2017 27-30 June 2017 – University of Modena and RE
Profactor GmbH
• Established in 1995
• Austria's No. 1 in applied production research
• 74 employees
• Research for Industry
FAIM2017 27-30 June 2017 – University of Modena and RE
Contents
Motivation
The XROB System
Human Robot Interaction– Cooperation
– Collaboration
Future Work
FAIM2017 27-30 June 2017 – University of Modena and RE
Introduction
Aging workforce in Europe
Consequential skill drain
Solution:
Cooperation between Human(s) andMachine(s)
• Goal: Increase the effectiveness and efficiency
• Requirement: Safe and natural workflow
Smart assembly line robots
(Image Sources: Eurostat; Yaskawa Motomon Robotics)
FAIM2017 27-30 June 2017 – University of Modena and RE
Forms of Human Robot Interaction
Shen Yi, München, 2015Thiemermann (PhD Thesis), Heimsheim, 2005.
FAIM2017 27-30 June 2017 – University of Modena and RE
Batch Size 1 - Requirements
Paradigm shift from mass production to mass customization
Handle a variety of different tasks, and be able to be reprogrammed fast by non-robot expert
The aim of reducing the amount of programming required by an non-expert using natural modes of communication is still an open topic
M. R. Pedersen et. al, 2016
FAIM2017 27-30 June 2017 – University of Modena and RE
Skill Based Learning
Skills instantiate action
Form the building blocks of the task
Intuitive object-centered robot abilities, which can easily be parameterized by a non-expert
• Focus on knowledge sharing between robots with similar capabilities
• Action – perception loop
RoboEarth 2011
FAIM2017 27-30 June 2017 – University of Modena and RE
XROB System
PerceptionSystem
Application Development
Cognitive Reasoning
System
Planning and Execution System
OMPL 2010RRT-Connect (ICRA 2000)
ReconstructMe, Profactor GmbHCANDELOR, Profactor GmbHAkkaladevi et. al, (ICRA 2016)Akkaladevi et. al, (CGVIS 2015)
Akkaladevi et. al, (ECAI, 2016)Akkaladevi et. al, (IHCI 2016)Akkaladevi et. al, (HRI 2017)
FAIM2017 27-30 June 2017 – University of Modena and RE
The XROB System - Video
FAIM2017 27-30 June 2017 – University of Modena and RE
XROB GUI - Video
FAIM2017 27-30 June 2017 – University of Modena and RE
XROB System – Skill Based Learning
• Set of skills form a task
• Tasks combines to form a recipe (for the process)
• In built services for sensor and actuators
• Parametrization online– Position
– Scan
– Reference
– Evaluation
FAIM2017 27-30 June 2017 – University of Modena and RE
Applications
• Human Robot Cooperation– Hand-guided
– Tangible Interfaces
• Human Robot Collaboration– Interactive learning
FAIM2017 27-30 June 2017 – University of Modena and RE
Human Robot Cooperation
FAIM2017 27-30 June 2017 – University of Modena and RE
Human Robot Cooperation
Human robot cooperation – where both agents simultaneously work on the same work piece
The AssistMe Project
https://www.profactor.at/.../projects/assistme/
FAIM2017 27-30 June 2017 – University of Modena and RE
Human Robot Cooperation – Video (1)
FAIM2017 27-30 June 2017 – University of Modena and RE
Human Robot Cooperation – Video (2)
FAIM2017 27-30 June 2017 – University of Modena and RE
Human Robot Cooperation – Video (3)
FAIM2017 27-30 June 2017 – University of Modena and RE
Comparison of two robotic systems
Duration(in minutes) of parametrization for different robotic systems
FAIM2017 27-30 June 2017 – University of Modena and RE
Tangible Interfaces - Video
FAIM2017 27-30 June 2017 – University of Modena and RE
Human Robot Collaboration
Human Robot Collaboration is where both agents perform coordinated actions on the same task
Project KoMoProd
• Creating a “common understanding” of the task
• Generation of recipes online
https://www.profactor.at/.../projects/komoprod/
FAIM2017 27-30 June 2017 – University of Modena and RE
Human Robot Collaboration - Video
FAIM2017 27-30 June 2017 – University of Modena and RE
Interactive Learning
FAIM2017 27-30 June 2017 – University of Modena and RE
Contributions
Easy intuitive programming for non-experts
Programming by demonstration features with help of GUI
• Tangible interfaces
• Interactive learning
Applicability to Human robot interactive assembly tasks with varying complexity
FAIM2017 27-30 June 2017 – University of Modena and RE
Future work
MMAssist_II - Assistive Systems for production in Human machine cooperation context (FFG, 858623), started on 01.05.2017 (3 Year Project) with 25 partners (10 Scientific partners, 15 Industrial partners)
Homepage: www.mmassist.at
FAIM2017 27-30 June 2017 – University of Modena and RE
Thank you
DI Sharath Chandra Akkaladevi
PROFACTOR GmbH
Im Stadtgut A2 | 4407 Steyr-Gleink | Austria
Tel. +43(7252) 885-325 | Fax +43(7252) 885-101
www.profactor.at
This work is funded by the projects KoMoProd (Austrian Ministry for Transport, Innovation and Technology), AssistMe (FFG, 848653), CompleteMe (FFG, 849441) and MMAssistII (FFG, 858623)
FAIM2017 27-30 June 2017 – University of Modena and RE
Appendix
FAIM2017 27-30 June 2017 – University of Modena and RE
References
E. Dean-Leon et.al, "Robotic technologies for fast deployment of industrial robot systems," in Industrial Electronics Society, IECON 2016-42nd Annual Conference of the IEEE, 2016. F. Steinmetz and R. Weitschat, "Skill parametrization approaches and skill architecture for human-robot interaction," in Automation Science and Engineering (CASE), 2016 IEEE International Conference on, 2016. M. R. Pedersen et. al, , "Robot skills for manufacturing: From concept to industrial deployment," Robotics and Computer-Integrated Manufacturing, vol. 37, pp. 282-291, 2016. D. Holz et. al, "A skill-based system for object perception and manipulation for automating kitting tasks," in Emerging Technologies & Factory Automation (ETFA), 2015 IEEE 20th Conference on, 2015.M. R. Pedersen and V. Krüger, "Automated planning of industrial logistics on a skill-equipped robot," in IROS 2015 workshop Task Planning for Intelligent Robots in Service and Manufacturing, Hamburg, Germany, 2015. W. K. H. Ko, Y. Wu, K. P. Tee and J. Buchli, "Towards industrial robot learning from demonstration," in Proceedings of the 3rd International Conference on Human-Agent Interaction, 2015. B. Finkemeyer, T. Kröger and F. M. Wahl, "Executing assembly tasks specified by manipulation primitive nets," Advanced Robotics, vol. 19, pp. 591-611, 2005. M. N. Nicolescu et.al, "Natural methods for robot task learning: Instructive demonstrations, generalization and practice," in Proceedings of the second international joint conference on Autonomous agents and multiagent systems, 2003. H. Mosemann and F. M. Wahl, "Automatic decomposition of planned assembly sequences into skill primitives," IEEE transactions on Robotics and Automation, vol. 17, pp. 709-718, 2001. H. Bruyninckx and J. De Schutter, "Specification of force-controlled actions in the" task frame formalism-a synthesis," IEEE transactions on robotics and automation, vol. 12, pp. 581-589, 1996.
FAIM2017 27-30 June 2017 – University of Modena and RE
Forms of Human Robot Interaction
Shen, Yi: System für die Mensch-Roboter-Koexistenz in der Fließmontage (Forschungsberichte / IWB 305), München, 2015.Thiemermann, Stefan: Direkte Mensch-Roboter-Kooperation in der Kleinteilemontage mit einem SCARA-Roboter. Univ., Diss.--Stuttgart, 2005 (IPA-IAO-Forschung und -Praxis 411), Heimsheim, 2005.
Human and robots perform individual
tasksPassively guided robot assistance
Human and Robot working on the same
work piece
Human and Robot working on the same
task
Close proximity Provide Physical Support
Synchronized action
Coordinated action
Coexistence Assistance Cooperation Collaboration
Forms of Human Robot Interactions in an Assembly Process
Process SpeedSafety Requirements
FAIM2017 27-30 June 2017 – University of Modena and RE
XROB System Contd...
Easy-to-use features that significantly speed up commissioning and make the operation more cost-efficient and flexible than common programming methods
The special software architecture allows easy and intuitive creation of processes and configuration of the components of a robot system via a single user interface
FAIM2017 27-30 June 2017 – University of Modena and RE
XROB
XROB platform for building human robot interactions in an intuitive way
Enables applications requiring customized patterns of human robot interactions
A flexible quality inspection system with intuitive configuration capabilities