An Energy-Efficient Parallel Algorithm for Real-Time Near-Optimal - PowerPoint PPT Presentation

An Energy-Efficient Parallel Algorithm for Real-Time Near-Optimal UAV Path Planning D. Palossi a , A. Marongiu ab , L. Benini ab B. Forsberg a , M. Furci b , R. Naldi b , L. Marconi b a ETH Zürich, b Univeristy of Bologna NVidia GTC17 – Munich, October 10 th - 12 th , 2017 - #23356 D. Palossi et al. | 13.06.2017 | 1

Introduction There are many applications for autonomous Unmanned Aerial Vehicles (UAVs)  Surveillance  Aerial Mapping  We focus on standard-size quadrotors Entertainment  Rescue Mission  Standard-size quadrotors (~50cm, few Kg, ~100W) → computational bound due to  weight/battery One of the fundamental functional blocks for autonomous UAVs is the path planner  D. Palossi et al. | 13.06.2017 | 2

Energy Efficiency Requirements Current standard-size UAV Next generation micro/nano-size UAV Current system Next Gen system Size [ ∅ , weight ] 50 cm / few Kg few cm / few g power budgets for Propellers Power Cons. hundreds of W few W / hundred mW pico-size UAV [1] Processing Device Class desktop CPU LP/ULP embedded Cognitive Skills fully autonomous If we want bring advanced cognitive skills of state-of-the-art systems into the next generation autonomous vehicles → energy efficient algorithms are key [1] Progress on "pico" air vehicles, R.J. Wood, B. Finio, M. Karpelson, K. Ma, N.O. Perez-Arancibia, P .S. Sreetharan, H. T anaka, and J.P . Whitney, Int. D. Palossi et al. | 13.06.2017 | 3 Symp. on Robotics Research (invited paper), Flagstafg, Az, Aug. 2011.

Energy Efficiency Requirements Current standard-size UAV Next generation We look into parallelism + near optimality micro/nano-size UAV as key solution to guarantee the energy requirements Current system Next Gen system Size [ ∅ , weight ] 50 cm / few Kg few cm / few g power budgets for Propellers Power Cons. hundreds of W few W / hundred mW pico-size UAV [1] Processing Device Class desktop CPU LP/ULP embedded Cognitive Skills fully autonomous If we want bring advanced cognitive skills of state-of-the-art systems into the next generation autonomous vehicles → energy efficient algorithms are key [1] Progress on "pico" air vehicles, R.J. Wood, B. Finio, M. Karpelson, K. Ma, N.O. Perez-Arancibia, P .S. Sreetharan, H. T anaka, and J.P . Whitney, Int. D. Palossi et al. | 13.06.2017 | 4 Symp. on Robotics Research (invited paper), Flagstafg, Az, Aug. 2011.

Outline Path Planning Application  Graph computation and exploration  Naive approximate and Atomic version  Profile-based version  Limitations of the Naive Approach  Experimental Evaluation  System Characterization  Experimental Results  The Predictable Execution Model (PREM)  D. Palossi et al. | 13.06.2017 | 5

Path Planning Application Path Planning: constantly updates the  route of the vehicle based on information sensed in real time selects the best path  (according to specific metrics) responsible for preventing  collisions with dynamic, unexpected obstacles the reactivity of the UAV  depends on the path planner response time D. Palossi et al. | 13.06.2017 | 6

Graph Computation Quadrotor Automaton [1] Represents the kinematic and the constraints of the robot [1] M. Furci, A. Paoli, and R. Naldi. A supervisory control strategy for robot-assisted search and rescue in hostile environments. In Emerging Technologies Factory Automation D. Palossi et al. | 13.06.2017 | 7 (ETFA), 2013 IEEE 18th Conference on, pages 1–4, Sept 2013.

Graph Computation Quadrotor Automaton [1] Map Automaton Represents the kinematic and the constraints of the robot [1] M. Furci, A. Paoli, and R. Naldi. A supervisory control strategy for robot-assisted search and rescue in hostile environments. In Emerging Technologies Factory Automation D. Palossi et al. | 13.06.2017 | 8 (ETFA), 2013 IEEE 18th Conference on, pages 1–4, Sept 2013.

Graph Computation Quadrotor Automaton [1] Map Automaton Represents the kinematic and the constraints of the robot Represents location, possible connection and its constraints: obstacles [1] M. Furci, A. Paoli, and R. Naldi. A supervisory control strategy for robot-assisted search and rescue in hostile environments. In Emerging Technologies Factory Automation D. Palossi et al. | 13.06.2017 | 9 (ETFA), 2013 IEEE 18th Conference on, pages 1–4, Sept 2013.

Graph Computation Quadrotor Automaton [1] Map Automaton Represents the kinematic and the constraints of the robot Represents location, possible connection and its constraints: obstacles [1] M. Furci, A. Paoli, and R. Naldi. A supervisory control strategy for robot-assisted search and rescue in hostile environments. In Emerging Technologies Factory Automation D. Palossi et al. | 13.06.2017 | 10 (ETFA), 2013 IEEE 18th Conference on, pages 1–4, Sept 2013.

Graph Computation Quadrotor Automaton [1] Map Automaton Represents the kinematic and the constraints of the robot Represents location, possible connection and its constraints: obstacles Sequence of movements: go_45 - go_45 - go_45 [1] M. Furci, A. Paoli, and R. Naldi. A supervisory control strategy for robot-assisted search and rescue in hostile environments. In Emerging Technologies Factory Automation D. Palossi et al. | 13.06.2017 | 11 (ETFA), 2013 IEEE 18th Conference on, pages 1–4, Sept 2013.

Graph Computation Quadrotor Automaton [1] Map Automaton Represents the kinematic and the constraints of the robot Represents location, possible connection and its constraints: obstacles Sequence of movements: go_45 - go_45 - go_45 [1] M. Furci, A. Paoli, and R. Naldi. A supervisory control strategy for robot-assisted search and rescue in hostile environments. In Emerging Technologies Factory Automation D. Palossi et al. | 13.06.2017 | 12 (ETFA), 2013 IEEE 18th Conference on, pages 1–4, Sept 2013.

Graph Computation Quadrotor Automaton [1] Map Automaton Represents the kinematic and the constraints of the robot Represents location, possible connection and its constraints: obstacles Sequence of movements: go_45 - go_45 - go_45 [1] M. Furci, A. Paoli, and R. Naldi. A supervisory control strategy for robot-assisted search and rescue in hostile environments. In Emerging Technologies Factory Automation D. Palossi et al. | 13.06.2017 | 13 (ETFA), 2013 IEEE 18th Conference on, pages 1–4, Sept 2013.

Graph Computation Quadrotor Automaton [1] Map Automaton Represents the kinematic and the constraints of the robot Represents location, possible connection and its constraints: obstacles Sequence of movements: go_45 - go_45 - go_45 [1] M. Furci, A. Paoli, and R. Naldi. A supervisory control strategy for robot-assisted search and rescue in hostile environments. In Emerging Technologies Factory Automation D. Palossi et al. | 13.06.2017 | 14 (ETFA), 2013 IEEE 18th Conference on, pages 1–4, Sept 2013.

Graph Computation Quadrotor Automaton [1] Map Automaton Represents the kinematic and the constraints of the robot Represents location, possible connection and its constraints: obstacles Sequence of movements: go_45 - go_45 - go_45 [1] M. Furci, A. Paoli, and R. Naldi. A supervisory control strategy for robot-assisted search and rescue in hostile environments. In Emerging Technologies Factory Automation D. Palossi et al. | 13.06.2017 | 15 (ETFA), 2013 IEEE 18th Conference on, pages 1–4, Sept 2013.

Graph Computation Quadrotor Automaton [1] Map Automaton Represents the kinematic and the constraints of the robot Represents location, possible connection and its constraints: obstacles Obstacle detected in 2-3 Sequence of movements: go_45 - go_45 - go_45 [1] M. Furci, A. Paoli, and R. Naldi. A supervisory control strategy for robot-assisted search and rescue in hostile environments. In Emerging Technologies Factory Automation D. Palossi et al. | 13.06.2017 | 16 (ETFA), 2013 IEEE 18th Conference on, pages 1–4, Sept 2013.

Graph Computation Quadrotor Automaton [1] Map Automaton Represents the kinematic and the constraints of the robot Represents location, possible connection and its constraints: obstacles Sequence of movements: go_0 - go_0 - go_90 - go_90 [1] M. Furci, A. Paoli, and R. Naldi. A supervisory control strategy for robot-assisted search and rescue in hostile environments. In Emerging Technologies Factory Automation D. Palossi et al. | 13.06.2017 | 17 (ETFA), 2013 IEEE 18th Conference on, pages 1–4, Sept 2013.

Graph Computation Quadrotor Automaton [1] Map Automaton Represents the kinematic and the constraints of the robot Represents location, possible connection and its constraints: obstacles Sequence of movements: go_0 - go_0 - go_90 - go_90 [1] M. Furci, A. Paoli, and R. Naldi. A supervisory control strategy for robot-assisted search and rescue in hostile environments. In Emerging Technologies Factory Automation D. Palossi et al. | 13.06.2017 | 18 (ETFA), 2013 IEEE 18th Conference on, pages 1–4, Sept 2013.

Graph Computation Quadrotor Automaton [1] Map Automaton Represents the kinematic and the constraints of the robot Represents location, possible connection and its constraints: obstacles Sequence of movements: go_0 - go_0 - go_90 - go_90 [1] M. Furci, A. Paoli, and R. Naldi. A supervisory control strategy for robot-assisted search and rescue in hostile environments. In Emerging Technologies Factory Automation D. Palossi et al. | 13.06.2017 | 19 (ETFA), 2013 IEEE 18th Conference on, pages 1–4, Sept 2013.