www.dlr.de Chart 1 RobOps Approaching a Holistic and Unified Interface Service Definition for Future Robotic Spacecraft Steffen Jaekel, Bernhard Brunner (1) Christian Laroque, Zoran Pjevic (2) Felix Flentge (3) Thomas Krueger, and André Schiele (4) (1) German Aerospace Center (DLR), Robotics and Mechatronics Center (2) OHB System AG (3) ESOC (4) ESTEC
www.dlr.de Chart 2 Introduction Space Robotics - Future and already deployed robot applications in space: Justin - In-space robotic assembly (ISRA): SSRMS, SPDM - EVA assistance: SSRMS, Robonaut, DLR s Justin, (small) satellites for inspection: SPHERES - Robotic exploration: MER s - On-orbit servicing (OOS) for prolonging lifetime of operational satellites, repair & refuel (RRM), extend or upgrade functionality (Hubble) - OOS for active debris removal from LEO or re-orbiting into graveyard orbit in GEO ROKVISS e.deorbit
www.dlr.de Chart 3 Challenges of Robotic Spacecraft Source: Airbus, DLR
www.dlr.de Chart 4 Distributed Mission Architecture - METERON
www.dlr.de Chart 5 Current Mission Operation Standards - Current Mission Operations standards e.g. Packet Utilization Standard (PUS) - are made for the operation of classic spacecraft mostly from one ground station - not for autonomous space robots Transport Semantics - Partly new CCSDS approaches to standardization: - File delivery, Asynchronous Message Service (AMS), Disruption-tolerant Network (DTN) - MO: Message Abstraction Layer (MAL), Monitor and Control Common Services (MO) - RobOps approach: - Effective control approach for autonomous robotic spacecraft, attached robotic devices and arbitrary subsystem equipment holistic - Clear distinction between interface service semantics and method of transport
www.dlr.de Chart 6 RobOps Study Contents 1. Communication 2. Control Modes 3. Roles & Responsibilities 4. Scenario Analysis 5. Service Definition 6. Technology Analysis & Implementation 7. Demonstration Prototype
www.dlr.de Chart 7 Communication - Dominant barrier: space - Properties of communication channel restricts capabilities - Higher delay, jitter, low data bandwidth increased autonomy - Communication window relay satellites - Deep-space communication with DTN
www.dlr.de Chart 8 Possible Control Modes Autonomous Control Algorithms pass real-time decisions based on a variety of sensor input and control the spacecraft and robotic manipulator No human control - Possible control modes depend on mission architecture and given communication delay as major barrier between operator and robot System Autonomy Shared/Supervized Autonomy/Control Autonomous algorithms execute human high-level control inputs, e.g. waypoints for path planning algorithms High-level human control Assisting Autonomy Autonomous algorithms support the operator in his manual system control, e.g. collision avoidance Mid-level human control Manual Control Interaction / Human Control - More system autonomy equals less human control and intervention - Manual control, assisting, shared and supervized autonomy - Monitoring: global, subsystem specific in real-time (telepresence), ad-hoc, post-hoc The human operator steers the system completely manual Low-level human control Telepresence with haptic feedback
www.dlr.de Chart 9 Roles and Responsibilities - Classic roles and responsibilities were analyzed for classic satellite operations - Partly confusing nomenclature differences between ESOC, GSOC and NASA - Robotic operations: - Very small time scale for reaction Robotic Spacecraft Satellite - Direct control of robotic payload and decisionmaking by robotic operator (RO) Remote site Orbiter RO astronaut-controlled robot Commander - Robotic Operations Manager (ROM) and Engineer (ROE) set goals and supervise operations RO Ground-controlled robot Mission control center command Robotic satellite operations SOM SOE SPACON Classic satellite operations report
www.dlr.de Chart 10 Scenario Analysis Use Cases Scenarios OOS Rover Free-Flyer Service List
www.dlr.de Chart 11 Structuring Telerobotic Services 1. Functions 1. 2. Configuration Monitoring Control 2. Mission Architecture 3. Levels of Autonomy Mission Mission Three Tier (3T ) layers Main service classes 1. Deliberative Layer Planning System System 1 System n New plan Scheduling Re-plan Subsystem Subsystem 1 Subsystem n 2. Sequencing Layer Action sense sense sense Precondition Postcondition Event Event System Level 1 System 1 3. Reactive Layer New action Action Action finished System Level 2 System 2 System 4 System Level 3 System 5 System 6
3T Layers of Autonomy www.dlr.de Chart 12 Architecture: (A) Mission 1:n (B) System 1:n (C) Subsystem Operator Function: (1) Control (2) Mon. (3) Config. (1) Control (2) Mon. (3) Config. (1) Control (2) Mon. (3) Config. Deliberative Layer Planning Mission Planning System Planning Sybsystem Planning (e.g. path planning) Sequencing Layer Mission Scheduling Scheduling MissionEvent System Scheduling Event System Event Subsystem Scheduling (e.g. path sequencing) Subaystem Event Reactive Layer Mission Action Action System Action Subsystem Action (privatizable) Common Services Monitoring Mission Monitoring System Monitoring Autonomous Operations Monitoring Path Subsystem Monitoring Classic Spacecraft Operations Control & Config. Path
www.dlr.de Chart 13 Service Privatizations - Functional approach to Monitoring & Control focus on autonomy rather than specific subsystem or device (PCDU, AOCS, Robot Arm) - Specific functionality is addresses via Privatizations - Privatizations become Subservices, to be used across missions Service, e.g. Action Service Operation, e.g. executeaction Operation specification 1. Subservice, e.g. Robotic 2. Type: Motion Parameters (defined by Subservice) e.g. cmd=setpose, device=arm, mode=cart, syntax=xyz_euler, reference=abs, posedata=(trans_x, trans_y, trans_z, rot_x, rot_y, rot_z)
www.dlr.de Chart 14 Technology Analysis - Overview - Message Abstraction Layer (MAL) and Mission Operations Services (MO) - Communications: - Message Based Communications - Data Distribution Service (DDS) - ActiveMQ - ømq (ZeroMQ) - Asynchronous Message Service (AMS) over DTN - File Based Communication - CCSDS File Delivery Protocol (CFDP) - File Transfer on Ground - Decision fell on MAL over DTN for transport RobOps MAL DTN
www.dlr.de Chart 15 Implementation - Overview
www.dlr.de Chart 16 High-Level Architecture of Robot Demonstrator
www.dlr.de Chart 17 Demonstration at ESTEC - Privatizations - Iterative implementation and testing approach - Picture: local testing at ESTEC Telerobotics and Haptics Laboratory with KUKA LWR-III setup and DTN over Intranet
www.dlr.de Chart 18
www.dlr.de Chart 19 Study Conclusions - Holistic control approach for autonomous robotic spacecraft - Action & Monitoring services were implemented and demonstrated with KUKA LWR and MOCUP rover from Telespazio - Possible future developments: Event service, artificial communication delay and disruptions in DTN link, complex scenario with control authority hand-over OOS Rover Free-Flyer
www.dlr.de Chart 20 The future of robots in space robotic exploration satellite servicing EVA support
www.dlr.de Chart 21 Use Cases and Mission Scenario Analysis - Use case analysis for space-robotic mission - What tasks have to be be done? uc Ov erv iew of General Use Cases Commissioning of Components «include» Spacecraft Operation of Heterogeneous Components Autonomous Operations - Layout of detailed scenarios as a composition of use cases for OOS, rover exploration and EVA support (free-flyer around ISS) in order to identify required service functionality Agent «invokes» FDIR «invokes» Mission Operations «extend» «include» «include» «include» «include» Supervised Autonomy Operations Shared Autonomy Operations «include» «invokes» Mission Briefing Manual Operations Mission Monitoring «include» Mission Data Logging
www.dlr.de Chart 22 Architecture of Remote Robot System
www.dlr.de Chart 23 Iterative Development and Demonstration KUKA LWR Robot Simulator & cmd GUI servicelayer.robot monitoring servicelayer.mc monitoring & action Port 4556 bidirectional On both sides DLR KUKA LWR Robot Simulator servicelayer.robot monitoring action servicelayer.mc Robot Viewer Command UI DLR KUKA LWR Robot Simulator UDP servicelayer.robot (server) rmc-thalia monitoring action servicelayer.mc (client) Robot Viewer Command UI DLR Telespazio Vega KUKA LWR Robot servicelayer.robot monitoring action servicelayer.mc Robot Viewer Command UI ESTEC KUKA LWR Robot servicelayer.robot monitoring action servicelayer.mc Robot Viewer Command UI ESTEC Telespazio Vega
www.dlr.de Chart 24 Demonstration at ESTEC - Demonstration of selected interface services: Action and Monitoring - Possible future developments: Event service, artificial communication delay and disruptions in DTN link, complex scenario with control authority hand-over