pith. sign in

arxiv: 1906.10513 · v1 · pith:PAM5CNO5new · submitted 2019-06-24 · 💻 cs.RO

The Role of Compute in Autonomous Aerial Vehicles

Behzad Boroujerdian , Hasan Genc , Srivatsan Krishnan , Bardienus Pieter Duisterhof , Brian Plancher , Kayvan Mansoorshahi , Marcelino Almeida , Wenzhi Cui
show 2 more authors
Aleksandra Faust Vijay Janapa Reddi
This is my paper

Pith reviewed 2026-05-25 17:52 UTC · model grok-4.3

classification 💻 cs.RO
keywords micro aerial vehiclescyber-physical co-designcompute subsystemmission timeenergy efficiencyMAVBenchautonomous robotsmotion planning
0
0 comments X

The pith

Compute and motion are tightly intertwined in micro aerial vehicles, requiring cyber-physical co-design for optimal mission time and energy.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper investigates how the compute subsystem affects mission time and energy in micro aerial vehicles, showing that compute decisions influence motion planning and control through multiple interconnected paths. It argues that traditional algorithm-focused improvements are insufficient and that hardware choices in computing must be considered alongside physical design. A sympathetic reader would care because this co-design approach could extend vehicle endurance for applications like package delivery. The authors support their case with analytical models, a closed-loop simulator, end-to-end benchmarks, and the open-sourced MAVBench toolset, concluding that separate development of cyber and physical elements falls short.

Core claim

Compute and motion are tightly intertwined in cyber-physical mobile machines such as MAVs, so a close examination of cyber and physical processes and their impact on one another is necessary. Different impact paths exist through which compute affects mission metrics, and these can be examined using analytical models, simulation, and benchmarking. Cyber-physical co-design, where robot cyber and physical processes are developed with mutual consideration, is required for optimal robot design.

What carries the argument

Cyber-physical co-design, the methodology of developing a robot's cyber and physical processes and quantities with mutual consideration, analogous to hardware-software co-design.

If this is right

  • Compute choices affect mission metrics through multiple paths including effects on control loop timing and motion planning.
  • Onboard compute improvements can address endurance limits beyond what algorithms alone achieve.
  • End-to-end benchmarking with closed-loop simulation is required to measure real impacts of compute on MAV performance.
  • MAVBench provides a shared toolset of simulator and benchmarks for studying these compute-motion interactions.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Design processes for other mobile robots could adopt early integration of compute hardware selection with physical parameters.
  • Trade-offs may exist where higher compute power draw is offset by shorter flight times or more efficient paths.
  • The methodology suggests testing whether co-design yields larger gains when scaled to multi-vehicle or longer-duration missions.

Load-bearing premise

The analytical models, closed-loop simulator, and benchmark suite accurately capture the dominant impact paths between compute subsystem choices and real-world MAV mission time and energy metrics.

What would settle it

A controlled test that varies only the compute hardware while holding physical design, algorithms, and environment fixed and measures no resulting change in mission time or energy consumption.

Figures

Figures reproduced from arXiv: 1906.10513 by Aleksandra Faust, Bardienus Pieter Duisterhof, Behzad Boroujerdian, Brian Plancher, Hasan Genc, Kayvan Mansoorshahi, Marcelino Almeida, Srivatsan Krishnan, Vijay Janapa Reddi, Wenzhi Cui.

Figure 1
Figure 1. Figure 1: Currently and predicted number of registered UAVs according to FAA [ [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: MAV robot complex. The three main subsystems, i.e., compute, sensors and actuators, of an [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Closed-loop data flow in a MAV. Information flows from sensors collecting environment data [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: MAVs based on battery capacity and size. Endurance is important for MAVs to be useful in the [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Cyber-physical interaction graph. This graph captures how the various subsystems of a robot [PITH_FULL_IMAGE:figures/full_fig_p008_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Cyber-physical interaction graph for our quadrotor MAV with some path examples. [PITH_FULL_IMAGE:figures/full_fig_p009_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Architectural overview of our closed-loop simulation. [PITH_FULL_IMAGE:figures/full_fig_p011_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: High-level application pipeline for a typical MAV application. The upper row presents a universal [PITH_FULL_IMAGE:figures/full_fig_p014_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: MAVBench workloads. Each workload is an end-to-end application targeting both industry and [PITH_FULL_IMAGE:figures/full_fig_p015_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Application dataflows. Circles and arrows denote nodes and their communications respectively. Sub￾scriber/publisher communication paradigm is denoted with filled black arrows whereas client/server with dotted red ones. Dotted black arrows denote various localization techniques. 16 [PITH_FULL_IMAGE:figures/full_fig_p016_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Three impact clusters, performance, mass, and power, impacting mission time and energy. [PITH_FULL_IMAGE:figures/full_fig_p020_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Theoretical max velocity and response time relationship. [PITH_FULL_IMAGE:figures/full_fig_p021_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: Obstacle avoidance in action, a bird’s-eye view. Note the progression in time as a result of [PITH_FULL_IMAGE:figures/full_fig_p022_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: Obstacle avoidance with the PPC pipeline. Latency associated with each stage is denoted [PITH_FULL_IMAGE:figures/full_fig_p023_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: Relationship between SLAM throughput (FPS) and maximum velocity and energy of UAVs. [PITH_FULL_IMAGE:figures/full_fig_p024_15.png] view at source ↗
Figure 16
Figure 16. Figure 16: Core/frequency sensitivity analysis of mission average velocity for various benchmarks. [PITH_FULL_IMAGE:figures/full_fig_p025_16.png] view at source ↗
Figure 17
Figure 17. Figure 17: Core/frequency sensitivity analysis of mission time for various benchmarks. [PITH_FULL_IMAGE:figures/full_fig_p025_17.png] view at source ↗
Figure 18
Figure 18. Figure 18: Kernel breakdown for MAVBench. The abbreviations are as follows: [PITH_FULL_IMAGE:figures/full_fig_p026_18.png] view at source ↗
Figure 19
Figure 19. Figure 19: Forces acting on a quadcoptor. Knowing these forces is necessary in understanding how [PITH_FULL_IMAGE:figures/full_fig_p027_19.png] view at source ↗
Figure 20
Figure 20. Figure 20: Impact of compute mass, a physical quantity, on velocity, acceleration and mission time. [PITH_FULL_IMAGE:figures/full_fig_p028_20.png] view at source ↗
Figure 21
Figure 21. Figure 21: Power collection. Drone is instrumented with eLogger and data is collected during flight. [PITH_FULL_IMAGE:figures/full_fig_p029_21.png] view at source ↗
Figure 22
Figure 22. Figure 22: Power profiling and breakdown. Compute generally makes up a small portion of the MAV’s [PITH_FULL_IMAGE:figures/full_fig_p029_22.png] view at source ↗
Figure 23
Figure 23. Figure 23: Core/frequency sensitivity analysis of mission energy for various benchmarks. [PITH_FULL_IMAGE:figures/full_fig_p030_23.png] view at source ↗
Figure 24
Figure 24. Figure 24: Impact of compute mass, a physical quantity, on power and energy. [PITH_FULL_IMAGE:figures/full_fig_p031_24.png] view at source ↗
Figure 25
Figure 25. Figure 25: Holistic impact of compute on mission metrics. The data enables cyber-physical co-design [PITH_FULL_IMAGE:figures/full_fig_p033_25.png] view at source ↗
Figure 26
Figure 26. Figure 26: Mission metrics design space with respect to different clusters. Mission metrics are shown as [PITH_FULL_IMAGE:figures/full_fig_p034_26.png] view at source ↗
Figure 27
Figure 27. Figure 27: Mission metrics’ gradient, the rate of change in the most optimal direction, and their com [PITH_FULL_IMAGE:figures/full_fig_p035_27.png] view at source ↗
Figure 28
Figure 28. Figure 28: Comparing a fully-onboard compute drone versus a fully-on-cloud drone. Our system allows [PITH_FULL_IMAGE:figures/full_fig_p036_28.png] view at source ↗
Figure 29
Figure 29. Figure 29: For the environment in (a), OctoMap’s resolution impact on the drone’s perception of its environment is shown in (b), (c), (d). Large resolution means larger voxel size (lower is better). We target the planning stage of the PPC pipeline and focus on the 3D Mapping as the application of choice to offload. As we show in [PITH_FULL_IMAGE:figures/full_fig_p037_29.png] view at source ↗
Figure 30
Figure 30. Figure 30: (a) Reduction in OctoMap resolution (accuracy) can be traded off with processing time. Increasing the x-axis means larger voxels to represent the space more coarsely (less accurately). A 6.5X reduction in resolution results in a 4.5X improvement in processing time. (b) Switching between OctoMap resolutions dynamically leads to successfully finishing the mission compared to 0.80 m. It also leads to battery… view at source ↗
read the original abstract

Autonomous-mobile cyber-physical machines are part of our future. Specifically, unmanned-aerial-vehicles have seen a resurgence in activity with use-cases such as package delivery. These systems face many challenges such as their low-endurance caused by limited onboard-energy, hence, improving the mission-time and energy are of importance. Such improvements traditionally are delivered through better algorithms. But our premise is that more powerful and efficient onboard-compute should also address the problem. This paper investigates how the compute subsystem, in a cyber-physical mobile machine, such as a Micro Aerial Vehicle, impacts mission-time and energy. Specifically, we pose the question as what is the role of computing for cyber-physical mobile robots? We show that compute and motion are tightly intertwined, hence a close examination of cyber and physical processes and their impact on one another is necessary. We show different impact paths through which compute impacts mission-metrics and examine them using analytical models, simulation, and end-to-end benchmarking. To enable similar studies, we open sourced MAVBench, our tool-set consisting of a closed-loop simulator and a benchmark suite. Our investigations show cyber-physical co-design, a methodology where robot's cyber and physical processes/quantities are developed with one another consideration, similar to hardware-software co-design, is necessary for optimal robot design.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 0 minor

Summary. The paper investigates the role of the onboard compute subsystem in Micro Aerial Vehicles (MAVs), arguing that compute choices significantly affect mission time and energy through multiple impact paths. Using analytical models, a closed-loop simulator (MAVBench), and end-to-end benchmarks, it shows that compute and motion are tightly coupled and concludes that cyber-physical co-design is necessary for optimal MAV design. The MAVBench toolset (simulator plus benchmark suite) is open-sourced to support similar studies.

Significance. If the models and simulator accurately represent dominant real-world couplings, the work usefully extends the design space for energy-constrained aerial robots beyond algorithmic improvements alone. The open-sourcing of MAVBench is a concrete strength that enables reproducibility and follow-on empirical work in the field.

major comments (1)
  1. [Abstract] Abstract: the central claim that cyber-physical co-design is necessary for optimal design depends on the assertion that the analytical models, MAVBench closed-loop simulator, and benchmarks capture the dominant impact paths between compute choices and mission time/energy. The manuscript provides no hardware validation of these models against physical MAVs; unmodeled effects (thermal throttling of sustained compute, sensor-to-actuator latency under payload variation, or vibration-induced drag) could therefore render the quantified impact paths artifacts rather than robust findings.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the constructive review. We address the major comment on hardware validation below.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the central claim that cyber-physical co-design is necessary for optimal design depends on the assertion that the analytical models, MAVBench closed-loop simulator, and benchmarks capture the dominant impact paths between compute choices and mission time/energy. The manuscript provides no hardware validation of these models against physical MAVs; unmodeled effects (thermal throttling of sustained compute, sensor-to-actuator latency under payload variation, or vibration-induced drag) could therefore render the quantified impact paths artifacts rather than robust findings.

    Authors: We agree that direct hardware validation on physical MAVs is absent and represents a limitation. The analytical models are derived from first-principles equations for MAV aerodynamics and compute energy, while MAVBench employs a closed-loop physics-based simulator (Gazebo) with parameters drawn from real MAV hardware datasheets and the benchmarks use realistic compute workloads. These elements demonstrate the compute-motion couplings within the modeled system. We will revise the manuscript to add an explicit limitations section discussing assumptions, the scope of the modeled impact paths, and unmodeled effects such as thermal throttling. This will clarify that our conclusions apply under the stated modeling assumptions and motivate the need for co-design even if additional real-world effects exist. The open-sourced toolset further enables such extensions by others. revision: partial

Circularity Check

0 steps flagged

No significant circularity; central claim rests on new empirical investigation and open-sourced simulator

full rationale

The paper derives its conclusion that cyber-physical co-design is necessary from analytical models, closed-loop simulation via the newly introduced MAVBench tool, and end-to-end benchmarks on MAV mission time/energy. No load-bearing step reduces by construction to a fitted parameter, self-citation chain, or renamed ansatz. The derivation chain is self-contained against external benchmarks and does not invoke uniqueness theorems or prior author results as the sole justification for its impact paths.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Based on abstract only; no explicit free parameters, axioms, or invented entities are identifiable from the provided text.

pith-pipeline@v0.9.0 · 5803 in / 1111 out tokens · 40261 ms · 2026-05-25T17:52:57.196676+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

63 extracted references · 63 canonical work pages · 3 internal anchors

  1. [1]

    [n. d.]. 3DR Robotics. https://3dr.com/solo-drone/

    work page

  2. [2]

    [n. d.]. Amazon delivered its first customer package by drone. https://www.usatoday.com/story/tech/news/2016/12/14/ amazon-delivered-its-first-customer-package-drone/95401366/

    work page arXiv 2016

  3. [3]

    [n. d.]. Eagle Tree Systems. http://www.eagletreesystems.com/index.php?route=product/product&product_id=54

    work page

  4. [4]

    [n. d.]. Embedded Systems Developer Kits, Modules, & SDKs | NVIDIA Jetson. https://www.nvidia.com/en-us/ autonomous-machines/embedded-systems-dev-kits-modules/

    work page

  5. [5]

    [n. d.]. FAA Aerospace Forecasts. https://www.faa.gov/data_research/aviation/aerospace_forecasts/

    work page

  6. [6]

    [n. d.]. Game Engine Technology by Unreal. https://www.unrealengine.com/en-US/what-is-unreal-engine-4

    work page

  7. [7]

    [n. d.]. IoT. https://software.intel.com/en-us/intel-joule-getting-started

    work page

  8. [8]

    [n. d.]. MA VLink: Micro Air Vehicle Protocol. https://github.com/mavlink

    work page

  9. [9]

    [n. d.]. Microsoft/AirSim: Open source simulator based on Unreal Engine for autonomous vehicles from Microsoft AI & Research. https://github.com/Microsoft/AirSim

    work page

  10. [10]

    [n. d.]. OMPL. https://ompl.kavrakilab.org/

    work page

  11. [11]

    [n. d.]. ORBSLAM 2. http://www.mujoco.org

    work page

  12. [12]

    [n. d.]. Physics Simulation | Unreal Engine. https://docs.unrealengine.com/latest/INT/Engine/Physics/

    work page

  13. [13]

    [n. d.]. Pixhawk Flight Controller. https://pixhawk.org/

    work page

  14. [14]

    [n. d.]. PX4 Architectural Overview - PX4 Developer Guide. https://dev.px4.io/en/concept/architecture.html

    work page

  15. [15]

    [n. d.]. ROS.org | Powering the world’s robots. http://www.ros.org/

    work page

  16. [16]

    Patrick Kwadwo Agyapong, Mikio Iwamura, Dirk Staehle, Wolfgang Kiess, and Anass Benjebbour. 2014. Design considerations for a 5G network architecture. IEEE Communications Magazine 52, 11 (2014), 65–75

    work page 2014

  17. [17]

    Althoff, M

    M. Althoff, M. Koschi, and S. Manzinger. 2017. CommonRoad: Composable benchmarks for motion planning on roads. In IEEE Intelligent Vehicles Symposium (IV). https://doi.org/10.1109/IVS.2017.7995802

    work page doi:10.1109/ivs.2017.7995802 2017

  18. [18]

    Arthur Holland Michel. [n. d.]. Amazon’s Drone Patents. http://dronecenter.bard.edu/amazon-drone-patents/

    work page

  19. [19]

    Bacha et al. 2008. Odin: Team victortango’s entry in the darpa urban challenge. Journal of field Robotics 25, 8 (2008), 467–492

    work page 2008

  20. [20]

    BBC News. [n. d.]. Google Plans Drone Delivery Service for 2017. http://www.bbc.com/news/technology-34704868

    work page 2017

  21. [21]

    Mehdi Benallegue, Jean-Paul Laumond, and Nicolas Mansard. 2017. Robot Motion Planning and Control: Is It More than a Technological Problem? 1–10 pages

    work page 2017

  22. [22]

    next-best-view

    Andreas Bircher, Mina Kamel, Kostas Alexis, Helen Oleynikova, and Roland Siegwart. 2016. Receding horizon "next-best-view" planner for 3D exploration. In IEEE International Conference on Robotics and Automation (ICRA). 40

    work page 2016

  23. [23]

    Min Chen and G. A. Rincon-Mora. 2006. Accurate electrical battery model capable of predicting runtime and I-V performance. in IEEE Transactions on Energy Conversion21, 2 (June 2006), 504–511. https://doi.org/10.1109/TEC. 2006.874229

    work page doi:10.1109/tec 2006

  24. [24]

    Anežka Chovancová, Tomáš Fico, ĽuboÅ¡ Chovanec, and Peter Hubinsk. 2014. Mathematical Modelling and Parameter Identification of Quadrotor (a survey). Procedia Engineering 96 (2014), 172 – 181. https://doi.org/10.1016/j. proeng.2014.12.139 Modelling of Mechanical and Mechatronic Systems

    work page doi:10.1016/j 2014

  25. [25]

    Navneet Dalal and Bill Triggs. 2005. Histograms of oriented gradients for human detection. In IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR)

    work page 2005

  26. [26]

    Cohen McCullough

    Debra R. Cohen McCullough. [n. d.]. Unmanned Aircraft Systems(UAS) Guidebook in Development. https://cops.usdoj. gov/html/dispatch/08-2014/UAS_Guidebook_in_Development.asp

    work page 2014

  27. [27]

    Ehredt, Dave. 2010. UAS: The Global Perspective. http://www.dcabr.org.br/download/eventos/eventos-realizados/2010/ seminario-vant-27-10-2010/cd-uvs-yearbook/pdf/P061-062_NATO_Dave-Ehredt.pdf

    work page 2010

  28. [28]

    Falanga, S

    D. Falanga, S. Kim, and D. Scaramuzza. 2019. How Fast Is Too Fast? The Role of Perception Latency in High-Speed Sense and Avoid. IEEE Robotics and Automation Letters 4, 2 (April 2019), 1884–1891. https://doi.org/10.1109/LRA. 2019.2898117

    work page doi:10.1109/lra 2019

  29. [29]

    Henriques

    Joao Faro, Christian Bailer, and Joao F. Henriques. 2013. Project Title. https://github.com/charlespwd/project-title

    work page 2013

  30. [30]

    C. D. Franco and G. Buttazzo. 2015. Energy-Aware Coverage Path Planning of UA Vs. InIEEE International Conference on Autonomous Robot Systems and Competitions (ICARSC). https://doi.org/10.1109/ICARSC.2015.17

    work page doi:10.1109/icarsc.2015.17 2015

  31. [31]

    Hasan Genc, Ting-Wu Chin, Matthew Halpern, and Vijay Janapa Reddi. 2016. Optimizing Sensor-Cloud Architecturesfor Real-time Autonomous Drone Operation

    work page 2016

  32. [32]

    Akhil Gupta and Rakesh Kumar Jha. 2015. A survey of 5G network: Architecture and emerging technologies. IEEE access 3 (2015), 1206–1232

    work page 2015

  33. [33]

    Peter E Hart, Nils J Nilsson, and Bertram Raphael. 1968. A formal basis for the heuristic determination of minimum cost paths. IEEE transactions on Systems Science and Cybernetics 4, 2 (1968), 100–107

    work page 1968

  34. [34]

    João F Henriques, Rui Caseiro, Pedro Martins, and Jorge Batista. 2015. High-speed tracking with kernelized correlation filters. IEEE Transactions on Pattern Analysis and Machine Intelligence37, 3 (2015)

    work page 2015

  35. [35]

    Gabriel Hoffmann, Haomiao Huang, Steven Waslander, and Claire J Tomlin. 2007. Quadrotor Helicopter Flight Dynamics and Control: Theory and Experiment. (08 2007). https://doi.org/10.2514/6.2007-6461

    work page doi:10.2514/6.2007-6461 2007

  36. [36]

    Armin Hornung, Kai M Wurm, Maren Bennewitz, Cyrill Stachniss, and Wolfram Burgard. 2013. OctoMap: An efficient probabilistic 3D mapping framework based on octrees. Autonomous Robots 34, 3 (2013), 189–206

    work page 2013

  37. [37]

    Dries Hulens, Jon Verbeke, and Toon Goedemé. 2016. Choosing the Best Embedded Processing Platform for On-Board UA V Image Processing. InComputer Vision, Imaging and Computer Graphics Theory and Applications, José Braz, Julien Pettré, Paul Richard, Andreas Kerren, Lars Linsen, Sebastiano Battiato, and Francisco Imai (Eds.). Springer International Publishin...

    work page 2016

  38. [38]

    James Rogers. [n. d.]. How drones are helping the Nepal earthquake relief effort. http://www.foxnews.com/tech/2015/ 04/30/how-drones-are-helping-nepal-earthquake-relief-effort.html

    work page 2015

  39. [39]

    Lydia E Kavraki, Petr Svestka, J-C Latombe, and Mark H Overmars. 1996. Probabilistic roadmaps for path planning in high-dimensional configuration spaces. IEEE transactions on Robotics and Automation 12, 4 (1996), 566–580

    work page 1996

  40. [40]

    N Koenig and A Howard. 2004. Design and use paradigms for Gazebo, an open-source multi-robot simulator. IEEE International Conference on Intelligent Robots and Systems (IROS) (2004)

    work page 2004

  41. [41]

    K. S. R. Kumar, V . V . Sastry, O. C. Sekhar, D. K. Mohanta, D. Rajesh, and M. P. C. Varma. 2016. Design and fabrication of coulomb counter for estimation of SOC of battery. In IEEE International Conference on Power Electronics, Drives and Energy Systems (PEDES). https://doi.org/10.1109/PEDES.2016.7914473

    work page doi:10.1109/pedes.2016.7914473 2016

  42. [42]

    Steven M LaValle. 1998. Rapidly-exploring random trees: A new tool for path planning. (1998)

    work page 1998

  43. [43]

    John Leonard, David Barrett, Jonathan How, Seth Teller, Matt Antone, Stefan Campbell, Alex Epstein, Gaston Fiore, Luke Fletcher, Emilio Frazzoli, et al. 2007. Team MIT urban challenge technical report. (2007)

    work page 2007

  44. [44]

    Sikang Liu, Michael Watterson, Sarah Tang, and Vijay Kumar. 2016. High speed navigation for quadrotors with limited onboard sensing. In IEEE International Conference on Robotics and Automation (ICRA)

    work page 2016

  45. [45]

    Matthias Mueller, Neil Smith, and Bernard Ghanem. 2016. A Benchmark and Simulator for UA V Tracking. InProc. of the European Conference on Computer Vision (ECCV)

    work page 2016

  46. [46]

    Raúl Mur-Artal and Juan D. Tardós. 2017. ORB-SLAM2: an Open-Source SLAM System for Monocular, Stereo and RGB-D Cameras. IEEE Transactions on Robotics 33, 5 (2017), 1255–1262. https://doi.org/10.1109/TRO.2017.2705103

    work page doi:10.1109/tro.2017.2705103 2017

  47. [47]

    Nardi et al. 2015. Introducing SLAMBench, a performance and accuracy benchmarking methodology for SLAM. In IEEE International Conference on Robotics and Automation (ICRA). IEEE

    work page 2015

  48. [48]

    Andri Qiantori, Agung Budi Sutiono, Hadi Hariyanto, Hirohiko Suwa, and Toshizumi Ohta. 2012. An Emergency Medical Communications System by Low Altitude Platform at the Early Stages of a Natural Disaster in Indonesia. J. Med. Syst. 36 (2012), 12. https://doi.org/10.1007/s10916-010-9444-9 41

    work page doi:10.1007/s10916-010-9444-9 2012

  49. [49]

    Tong Qin, Peiliang Li, and Shaojie Shen. 2017. Vins-mono: A robust and versatile monocular visual-inertial state estimator. arXiv preprint arXiv:1708.03852 (2017)

    work page internal anchor Pith review Pith/arXiv arXiv 2017

  50. [50]

    Weichao Qiu and Alan Yuille. [n. d.]. Unrealcv: Connecting computer vision to unreal engine. InComputer Vision–ECCV 2016 Workshops. Springer

    work page 2016

  51. [51]

    Rachel Feltman. [n. d.]. The Future of Sports Photography:Drones. https://www.theatlantic.com/technology/archive/ 2014/02/the-future-of-sports-photography-drones/283896/

    work page 2014

  52. [52]

    Joseph Redmon and Ali Farhadi. 2016. YOLO9000: Better, Faster, Stronger. CoRR abs/1612.08242 (2016). arXiv:1612.08242 http://arxiv.org/abs/1612.08242

    work page internal anchor Pith review Pith/arXiv arXiv 2016

  53. [53]

    Francesco Sabatino. 2015. Quadrotor control: modeling, nonlinearcontrol design, and simulation

    work page 2015

  54. [54]

    Sayre-McCord et al. 2018. Visual-inertial navigation algorithm development using photorealistic camera simulation in the loop. In IEEE International Conference on Robotics and Automation (ICRA)

    work page 2018

  55. [55]

    Shital Shah, Debadeepta Dey, Chris Lovett, and Ashish Kapoor. 2017. AirSim: High-Fidelity Visual and Physical Simulation for Autonomous Vehicles. CoRR abs/1705.05065 (2017). arXiv:1705.05065 http://arxiv.org/abs/1705.05065

    work page internal anchor Pith review Pith/arXiv arXiv 2017

  56. [56]

    Bruno Siciliano and Oussama Khatib. 2007. Springer Handbook of Robotics. Springer-Verlag New York, Inc., Secaucus, NJ, USA

    work page 2007

  57. [57]

    Takafumi Taketomi, Hideaki Uchiyama, and Sei Ikeda. 2017. Visual SLAM algorithms: a survey from 2010 to 2016.IPSJ Transactions on Computer Vision and Applications9, 1 (02 Jun 2017), 16. https://doi.org/10.1186/s41074-017-0027-2

    work page doi:10.1186/s41074-017-0027-2 2017

  58. [58]

    Elbassioni, and Majid Khonji

    Chien-Ming Tseng, Chi-Kin Chau, Khaled M. Elbassioni, and Majid Khonji. 2017. Flight Tour Planning with Recharging Optimization for Battery-operated Autonomous Drones. CoRR abs/1703.10049 (2017). arXiv:1703.10049 http: //arxiv.org/abs/1703.10049

    work page arXiv 2017

  59. [59]

    Urmson et al. 2009. Autonomous driving in traffic: Boss and the urban challenge. AI magazine 30, 2 (2009), 17

    work page 2009

  60. [60]

    Augusto Vega, Pradip Bose, and Alper Buyuktosunoglu. 2017. Reliable electrical systems for micro aerial vehicles and insect-scale robots: Challenges and progress. In Rugged Embedded Systems. Morgan Kaufmann. https://doi.org/10. 1016/B978-0-12-802459-1.00007-5

    work page 2017

  61. [61]

    Watts, Vincent G

    Adam C. Watts, Vincent G. Ambrosia, and Everett A. Hinkley. 2012. Unmanned Aircraft Systems in Remote Sensing and Scientific Research: Classification and Considerations of Use. Remote Sensing 4, 6 (2012), 1671–1692. https: //doi.org/10.3390/rs4061671

    work page doi:10.3390/rs4061671 2012

  62. [62]

    Weisz, Y

    J. Weisz, Y . Huang, F. Lier, S. Sethumadhavan, and P. Allen. 2016. RoboBench: Towards sustainable robotics system benchmarking. In 2016 IEEE International Conference on Robotics and Automation (ICRA) . 3383–3389. https: //doi.org/10.1109/ICRA.2016.7487514

    work page doi:10.1109/icra.2016.7487514 2016

  63. [63]

    Xioji Chen. [n. d.]. Engineering Uber’s Self-Driving Car Visualization Platform for Web. https://eng.uber.com/ atg-dataviz/. 42

    work page