AI Planner Applications Practical Applications of AI Planners

AI Planner Applications Practical Applications of AI Planners AI Planner Applications

Overview l l l Deep Space 1 Other Practical Applications of AI Planners Common Themes AI Planner Applications 2

Literature l Deep Space 1 Papers l Ghallab, M. , Nau, D. and Traverso, P. , Automated Planning – Theory and Practice, chapter 19, . Elsevier/Morgan Kaufmann, 2004. Bernard, D. E. , Dorais, G. A. , Fry, C. , Gamble Jr. , E. B. , Kanfesky, B. , Kurien, J. , Millar, W. , Muscettola, N. , Nayak, P. P. , Pell, B. , Rajan, K. , Rouquette, N. , Smith, B. , and Williams, B. C. Design of the Remote Agent experiment for spacecraft autonomy. Procs. of the IEEEAerospace Conf. , Snowmass, CO, 1998. l l http: //nmp. jpl. nasa. gov/ds 1/papers. html l Other Practical Planners l l Ghallab, M. , Nau, D. and Traverso, P. , Automated Planning – Theory and Practice, chapter 22 and 23. Elsevier/Morgan Kaufmann, 2004 Tate, A. and Dalton, J. (2003) O-Plan: a Common Lisp Planning Web Service, invited paper, in Proceedings of the International Lisp Conference 2003, October 12 -25, 2003, New York, NY, USA, October 12 -15, 2003. l http: //www. aiai. ed. ac. uk/project/ix/documents/2003 -luc-tate-oplan-web. doc AI Planner Applications 3

Deep Space 1 – 1998 -2001 http: //nmp. jpl. nasa. gov/ds 1/ AI Planner Applications 4

DS 1 – Comet Borrelly http: //nmp. jpl. nasa. gov/ds 1/ AI Planner Applications 5

DS 1 Domain Requirements Achieve diverse goals on real spacecraft l High Reliability • • l l single point failures multiple sequential failures Tight resource constraints • • resource contention conflicting goals Hard-time deadlines Limited Observability Concurrent Activity AI Planner Applications 6

DS 1 Remote Agent Approach l l l Constraint-based planning and scheduling • Robust multi-threaded execution • supports reliability, concurrency, deadlines Model-based fault diagnosis and reconfiguration • l supports goal achievement, resource constraints, deadlines, concurrency supports limited observability, reliability, concurrency Real-time control and monitoring AI Planner Applications 7

DS 1 Levels of Autonomy Listed from least to most autonomous mode: 1. single low-level real-time command execution 2. time-stamped command sequence execution 3. single goal achievement with auto-recovery 4. model-based state estimation & error detection 5. scripted plan with dynamic task decomposition 6. on-board back-to-back plan generation, execution, & plan recovery AI Planner Applications 8

DS 1 Levels of Autonomy AI Planner Applications 9

DS 1 Systems Planning Execution Monitoring AI Planner Applications 10

DS 1 RAX Functionality PS/MM l generate plans on-board the spacecraft l reject low-priority unachievable goals l replan following a simulated failure l enable modification of mission goals from ground EXEC l provide a low-level commanding interface l initiate on-board planning l execute plans generated both on-board and on the ground l recognize and respond to plan failure l maintain required properties in the face of failures MIR l confirm executive command execution l demonstrate model-based failure detection, isolation, and recovery l demonstrate ability to update on-board state via ground commands AI Planner Applications 11

DS 1 Remote Agent (RA) Architecture AI Planner Applications 12

DS 1 Planner Architecture AI Planner Applications 13

DS 1 Diversity of Goals l l l Final state goals • “Turn off the camera once you are done using it” Scheduled goals • “Communicate to Earth at pre-specified times” Periodic goals • “Take asteroid pictures for navigation every 2 days for 2 hours” Information-seeking goals • “Ask the on-board navigation system for the thrusting profile” Continuous accumulation goals • “Accumulate thrust with a 90% duty cycle” Default goals • “When you have nothing else to do, point HGA to Earth” AI Planner Applications 14

DS 1 Diversity of Constraints l l l State/action constraints • “To take a picture, the camera must be on. ” Finite resources • power True parallelism • the ACS loops must work in parallel with the IPS controller Functional dependencies • “The duration of a turn depends on its source and destination. ” Continuously varying parameters • amount of accumulated thrust Other software modules as specialized planners • on-board navigator AI Planner Applications 15

DS 1 Domain Description Language AI Planner Applications Temporal Constraints in DDL Command to EXEC in ESL 16

DS 1 Plan Fragment AI Planner Applications 17

DS 1 RA Exec Status Tool AI Planner Applications 18

DS 1 RA Ground Tools AI Planner Applications 19

DS 1 – Flight Experiments 17 th – 21 st 1999 l l l RAX was activated and controlled the spacecraft autonomously. Some issues and alarms did arise: Divergence of model predicted values of state of Ion Propulsion System (IPS) and observed values – due to infrequency of real monitor updates. EXEC deadlocked in use. Problem diagnosed and fix designed by not uploaded to DS 1 for fears of safety of flight systems. Condition had not appeared in thousands of ground tests indicating needs formal verification methods for this type of safety/mission critical software. Following other experiments, RAX was deemed to have achieved its aims and objectives. AI Planner Applications 20

DS 1 Experiment 2 Day Scenario AI Planner Applications 21

DS 1 Summary Objectives and Capabilities AI Planner Applications 22

Earlier Spacecraft Planning Applications l l l l l AI Planner Applications Deviser NASA Jet Propulsion Lab Steven Vere, JPL First NASA AI Planner 1982 -3 Based on Tate’s Nonlin Added Time Windows Voyager Mission Plans Not used live 23

Earlier Spacecraft Planning Applications l l l AI Planner Applications T-SCHED Brian Drabble, AIAI BNSC T-SAT Project 1989 Ground-based plan generation 24 hour plan uploaded and executed on Uo. SAT-II 24

Some Other Practical Applications of AI Planning l l l l Nonlin electricity generation turbine overhaul Deviser Voyager mission planning demonstration SIPE – a planner that can organise a …. brewery Optimum-AIV • • Integrating technologies Integrating with other IT systems O-Plan various uses – see next slides Bridge Baron Deep Space 1 – to boldly go… AI Planner Applications 25

Practical Applications of AI Planning – O-Plan Applications O-Plan has been used in a variety of realistic applications: l Noncombatant Evacuation Operations (Tate, et al. , 2000 b) l Search & Rescue Coordination (Kingston et al. , 1996) l US Army Hostage Rescue (Tate et al. , 2000 a) l Spacecraft Mission Planning (Drabble et al. , 1997) l Construction Planning (Currie and Tate, 1991 and others) l Engineering Tasks (Tate, 1997) l Biological Pathway Discovery (Khan et al. , 2003) l Unmanned Autonomous Vehicle Command Control l O-Plan’s design was also used as the basis for Optimum-AIV (Arup et al. , 1994), a deployed system used for assembly, integration and verification in preparation of the payload bay for flights of the European Space Agency Ariane IV launcher. AI Planner Applications 26

Practical Applications of AI Planning – O-Plan Features A wide variety of AI planning features are included in O-Plan: l Domain knowledge elicitation l Rich plan representation and use l Hierarchical Task Network Planning l Detailed constraint management l Goal structure-based plan monitoring l Dynamic issue handling l Plan repair in low and high tempo situations l Interfaces for users with different roles l Management of planning and execution workflow AI Planner Applications 27

Common Themes in Practical Applications of AI Planning l l l Outer HTN “human-relatable” approach Underlying rich time and resource constraint handling Integration with plan execution Model-based simulation and monitoring Rich knowledge modelling languages and interfaces AI Planner Applications 28

Summary l l l Deep Space 1 and Remote Agent Experiment Other Practical Applications of AI Planners Common Themes AI Planner Applications 29

Literature l Deep Space 1 Papers l Ghallab, M. , Nau, D. and Traverso, P. , Automated Planning – Theory and Practice, chapter 19, . Elsevier/Morgan Kaufmann, 2004. Bernard, D. E. , Dorais, G. A. , Fry, C. , Gamble Jr. , E. B. , Kanfesky, B. , Kurien, J. , Millar, W. , Muscettola, N. , Nayak, P. P. , Pell, B. , Rajan, K. , Rouquette, N. , Smith, B. , and Williams, B. C. Design of the Remote Agent experiment for spacecraft autonomy. Procs. of the IEEEAerospace Conf. , Snowmass, CO, 1998. l l http: //nmp. jpl. nasa. gov/ds 1/papers. html l Other Practical Planners l l Ghallab, M. , Nau, D. and Traverso, P. , Automated Planning – Theory and Practice, chapter 22 and 23. Elsevier/Morgan Kaufmann, 2004 Tate, A. and Dalton, J. (2003) O-Plan: a Common Lisp Planning Web Service, invited paper, in Proceedings of the International Lisp Conference 2003, October 12 -25, 2003, New York, NY, USA, October 12 -15, 2003. l http: //www. aiai. ed. ac. uk/project/ix/documents/2003 -luc-tate-oplan-web. pdf AI Planner Applications 30