1 / 43

Standards or no Standards in Advanced Robotics that is the Question Erwin Prassler

Standards or no Standards in Advanced Robotics that is the Question Erwin Prassler B-IT Bonn-Aachen Int. Center for Information Technology Applied Science Institute Grantham-Allee 20, 53757 St. Augustin, GERMANY erwin.prassler@fh-brs.de. Standards …. … or no standards in car manufacturing.

dougal
Download Presentation

Standards or no Standards in Advanced Robotics that is the Question Erwin Prassler

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Standards or no Standards in Advanced Robotics that is the Question Erwin Prassler B-IT Bonn-Aachen Int. Center for Information TechnologyApplied Science Institute Grantham-Allee 20, 53757 St. Augustin, GERMANY erwin.prassler@fh-brs.de

  2. Standards …

  3. … or no standards in car manufacturing

  4. No standards … CAD-model of first prototype

  5. … or standards in robotics design Challenges • integration of heterogeneous modules • interchangeability of modules • integration of hybrid control paradigms • coordination of a multitude of technical components and their behaviors • robustness • self-modeling, monitoring, diagnosis • maintainability • re-usability CAD-model of first prototype

  6. “Commonalities” in the robotics domain • Extremely heterogeneous hardware • Inherently concurrent • Inherently distributed • Device dependent • Stochastic properties of physical world • Real-time constrained • Resource constrained • Currently not adequately supported by available • robot software architectures • robot software development environments • Inadequate evaluation and assessment • Mere demonstration character “Commonalities” in the robotics domain

  7. Design of Autonomous RobotsHeterogeneity of Hardware • Robots, robot teams, sensor networks are distributed system composed of very heterogeneous hardware • sensors: • bumpers, IRs, sonars, laser scanners, accelerometers, gyros, GPS, microphones, cameras, omni-cams, stereo-heads • actuators: • DC motors, steppers, servos, kickers, pan-tilts, arms, hands, legs, HDoF bodies, polymorphic systems • computational entities: • microcontrollers, embedded PCs, PDAs, notebooks, remote PCs • communication devices, mechanisms, and protocols: • I2C, serial, CAN, USB, UDP, TCP/IP, Firewire • No plug and play! • No configuration management! • Heterogeneity grows over system lifetime! “Commonalities” in the robotics domain

  8. Design of Autonomous RobotsDiversity of Software • Roboticists use a wide diversity of often computationally intensive methods • Control theory • Computational geometry • Neural networks • Genetic algorithms and evolutionary methods • Reinforcement learning • Vision processing routines • AI planning techniques • Behavior systems • Probabilistic reasoning • Optimization techniques • Search techniques • All these problems make software development for mobile robots very complex and error-prone “Commonalities” in the robotics domain

  9. B-IT Tutorial on Robot Middleware and Integration Frameworks • Player/Stage/Gazebo • MCA2 • Smartsoft • Miro • Marie • ORCA2 “Commonalities” in the robotics domain

  10. Service RobotApplications VacuumBot NurseBot ShopBot NannyBot VacuumBot NurseBot ShopBot NannyBot VacuumBot0 NurseBot0 ShopBot0 NannyBot0 - application frameworks RCAFramework Layer VacuumBot NurseBot ShopBot NannyBot + domain knowledge - component libraries Robot ComponentFramework Layer SmartSoft Builder ORCA Components X Component + CBSE - functional class libraries Robot Method Framework Layer Miro LAP RHI GUI Miro VIP + methods + patterns + generic utilities Miro MCL Vision Object Recog Object Track SLAM Path Planning Task Planning Learning Logging Miro BAP Marie MediatorPatterns Player Fiducial SmartsoftComm Patterns MCA2 Control Miro Logging - services Network Service Layer Arm Service WebServices Laser Service + network-transp. access Base Drive Service CORBA Services - classes Class Layer ArmClass LaserClass + object-orientation BaseDriveClass CommClass - file I/O File Interface Layer ArmFileIF LaserFileIF BaseDriveFileIF + coherent file IF CommFileIF - functions + protocols Device Driver Layer Arm Device Laser Device Comm Device + plurality of vendor IFs Base Drive Device “Commonalities” in the robotics domain

  11. All in all What Makes The Problem Hard? • no common architectures • no common methods • hardware-dependency of developed code • missing abstractions • no reusable components

  12. Standards or no Standards in Advanced Robotics is that really a Question?

  13. RoSta Robot Standards and Reference Architectures for Service Robots and Mobile Manipulation EU Coordination Action IST-045304

  14. www.robotics-platform.eu.com RoSta Robot Standards and Reference Architectures www.euron.org www.eu-nited-robotics.net www.service-robotik-initiative.de

  15. RoSta’s Overall Mission • Europe as key player in the definition of formal standards and the establishment of “de facto” standards in the field of robotics, especially service robotics. • Formulation of action plans for defining standards in a very few, selected key topics which have the highest possible impact. • Form the root of a whole chain of standard defining activities going beyond the specific activities of RoSta. • Topics • Glossary/ontology for mobile manipulation, service robots • Specification of a reference architecture • Specification of a middleware • Formulation of benchmarks

  16. Project Profile • Relation: FP6, Priority 2: “Information Society Technologies”,6th Call, 2.6.1 Advanced Robotics; Contract IST-045304 • Duration: Jan. 1st, 2007 to Dec. 31st, 2008 (24 months) • Budget: ~ 1 MEUR • Project Lead: Fraunhofer IPA • Project office: GPS Stuttgart

  17. Ultimate RoSta Deliverables • Each line of activity will result either in: • An action plan for a standard defining activity or • An action plan and a recommendation/proposal to the European Commission for a supported activity (e.g. a open-source project with financial support in FP7) or • An action plan for a community driven open-source activity with seed-money for example to run a project office or alike “Challenge 2: Cognitive Systems, Interaction, Robotics” (~€195m), future calls

  18. RoSta Overall Structure

  19. The RoSta Workplan WP i.1 State of the art • T0+1 Expert meeting (6-8 experts) • T0+3 Expert meeting dito • T0+5 Expert meeting dito • T0+7 Consolidation meetingwith all stakeholders WP i.2 Requirement analysis • T0+7 Expert meeting (6-8 experts) • T0+9 Expert meeting dito • T0+11 Expert meeting dito • T0+13 Consolidation meetingwith all stakeholders WP i.3 Action plan and recommendation • T0+13 Expert meeting 6 - 8 experts • T0+15 Expert meeting dito • T0+17 Expert meeting dito • T0+19 Consolidation meeting with all stakeholders T0+20: compilation of reports etc. Consolidationmeetings Expertmeetings RoSta Expert studies Expertgroup Stakeholders; Research, EC

  20. Cooperations with IEEE & OMG Consolidationmeetings Expertmeetings RoSta Expert studies Expertgroup Stakeholders; Research, EC

  21. RoSta Overall Time Line

  22. RoSta Overall Structure

  23. Robotsuppliers Standardizationbodies all Public EUnited IPAGPS Robot componentmanufacturers RoSta EC EURON Research,academia Service roboticsindustries Industrialstakeholders Selection of Experts Criteria: • Unbiased • Reliable • Key-players Two-step selection process: • Public announcement and CFP • Select 8-10 experts/activity Expert participation: • Contribution to RoSta Wiki • Attending at expert meetingsand work shops

  24. Middleware:Objectives • Objectives • Given the variety (“zoo”) of middleware and integration frameworks in robotics • identify the requirements of the academic and industrial community for communication infrastructure and distributed system design are met by available solutions • identify possible mismatches and unmet requirements • elaborate action plan for harmonizing, revising, expanding existing approaches

  25. Middleware:Tasks • Tasks • Compilation and evaluation of State of the Art in Robot Middleware • Requirement analysis • Action plan and recommendation

  26. Middleware:Wiki • RoSta wiki and mailing list on middleware and architecture • wiki.robot-standards.org/index.php/Middleware • middleware@robot-standards.org • considerable data on state of the art in bibliography and reports • currently app. 30 subscribers to RoSta wiki • standardization related activities/ideas, determination of existing standards and software which can be re-used. • new section on state of the art in robustness • state of the art on dependability in robotics (fault tolerance, robustness, etc.)

  27. Middleware:Wiki Wiki content • State of the art • Comparison and evaluation • Requirement analysis • Definitions/references • Dependability • Other robotic projects

  28. Middleware:Wiki State of the art content and structure

  29. Middleware:Wiki Sample description • System organization • Communication app. • Other features, fault tolerance

  30. Middleware:Wiki Section “Architectural practices”

  31. Middleware:Wiki Section “Re-use and Standardization”

  32. Middleware:Wiki Section “Re-use and Standardization”

  33. Middleware:Wiki Section “Bibliography”

  34. Middleware:Wiki Section “Catalogue of robotics software projects”

  35. Middleware:Wiki Section “Dependability in robotics”

  36. Middleware:Wiki Section “robotic fault tolerance and robustness”

  37. Middleware:Wiki • Expert meetings • expert meeting and a workshop at two biggest robotics conferences: ICRA 2007 and IROS 2008 • presentation of RoSta at OMG robotics DTF meeting • regular meetings with experts on the phone, mailing list and face-to-face (in cooperation with University of Lund and University of Leuven) • RoSta middleware questionnaire with EUnited is distributed to industrial community

  38. Middleware:Wiki Expert meetings Middleware for mobile manipulation and service robots

  39. Middleware:Wiki List of experts Anthony Mallet and Sara Fleury, LAAS RIA William D. Smart, Washington University in St. Louis Davide Brugali, University of Bergamo Brian Gerkey, SRI International Issa Nesnas, Hans Utz,NASA JPL Carle Cote, University of Sherbrooke Christoph Borst, DLR Robotics Herman Bruyninckx, University of Leuven Berthold Baeuml, German Aerospace Center Abheek Bose, ADA Robotics Chief Robotics Engineer Andrew Tanenbaum, Vrije University Alexey Makarenko, Alex Brooks and Tobias Kaup, University of Sydney Gerhard Kraetzschmar and Paul Ploeger, Fraunhofer IAIS

  40. Middleware:Wiki RoSta workshop at IROS 2007

  41. Middleware:Wiki Middleware/architecture workshop at IROS 2007

  42. Middleware:Wiki Questionnaire for requirement analysis

  43. Middleware:Lessons • Painful lessons: • Senior experts are totally overbooked. It is impossible to recruit them for a series of workshops in two months intervals • Contingency measure: • do as much work as possible by BSCW tools (Wiki), email, and VoIP resulted in a very comprehensive Wiki with a steadily increasing contributions by the community • transfer RoSta Wiki on Middleware to Wikipedia to increase reach-out • Efforts which need to be invested in the three tasks are NOT uniformly distributed. State of the Art can be done locally (personnel effort, little travel)Requirement analysis is very hard (higher personnel effort, significantly more travel • Contingency measure: • money not spend for Task 3.1 will go into Task 3.2 Middleware for mobile manipulation and service robots 2

More Related