ROScopter: A Multirotor Autopilot based on ROSflight 2.0

Jacob Moore; Ian Reid; Phil Tokumaru; Randy Beard; Tim McLain

cs.ROMar 5, 2026

ROScopter: A Multirotor Autopilot based on ROSflight 2.0

Authors: Jacob Moore, Ian Reid, Phil Tokumaru, Randy Beard, Tim McLain

Organizations: Brigham Young University · AeroVironment Inc.

Abstract

ROScopter is a lean multirotor autopilot built for researchers. ROScopter seeks to accelerate simulation and hardware testing of research code with an architecture that is both easy to understand and simple to modify. ROScopter is designed to interface with ROSflight 2.0 and runs entirely on an onboard flight computer, leveraging the features of ROS 2 to improve modularity. This work describes the architecture of ROScopter and how it can be used to test application code in both simulated and hardware environments. Hardware results of the default ROScopter behavior are presented, showing that ROScopter achieves similar performance to another state-of-the-art autopilot for basic waypoint-following maneuvers, but with a significantly reduced and more modular code-base.

Figures & tables

Fig. 1 : Architecture of ROScopter within the ROSflight framework. ROScopter communicates with the ROSflight firmware via the ROSflightIO node.

Fig. 2 : Entry points for the ROScopter controller. Each entry point is a PID controller and is described in Table I . The ROScopter controller sends commands to the ROSflight firmware.

	Name	Description of reference commands
0	NED-Pos Yaw	Inertial NED positions and yaw
1	NE-Vel D-Pos YawR	Inertial N-E velocities, D position, and yaw rate
2	FRD-Accel YawR	Vehicle-1 (front-right-down) accelerations and yaw rate
3	NED-Vel YawR	Inertial NED velocities and yaw rate
4	NE-Pos D-Vel Yaw	Inertial N-E velocities and yaw
5	Roll Pitch Yaw Throttle	Roll, pitch, yaw, and throttle

TABLE I : Description of each entry point into the ROScopter cascading controller architecture.

Waypoint	$w_{1}$	$w_{2}$	$w_{3}$
North (m)	0	-20	-20
East (m)	0	0	20
Down (m)	-5	-8	-5
Heading (deg)	130	130	130

TABLE II : Desired NED positions and headings of the waypoint mission shown in Figure 3 .

Fig. 3 : Top-down view of simulation and hardware flight test results using the exact same control and estimation gains on a HolyBro x650 quadcopter frame. Simulation and hardware responses match, highlighting how ROScopter can accelerate simulation to hardware transitions.

Fig. 4 : Response of the vehicle using ROScopter to commanded position and heading setpoints in hardware for the same trajectory shown in Figure 3 .

Fig. 5 : Errors between truth and estimated states for ROScopter’s estimator from simulation.

	X (m)	Y (m)	Z (m)	Total (m)
ROScopter	0.197	0.145	0.474	0.533
PX4	0.152	0.102	0.431	0.469

TABLE III : RMS errors between estimated state and desired straight-line waypoint path from hardware flight tests. PX4’s RMSE only includes data after

t=63

s, after altitude control had stabilized.

Fig. 6 : Response of the vehicle using PX4 in hardware when flying the same waypoint path shown in Figure 3 .

Fig. 7 : Comparison of the PX4 and ROScopter estimator output given the same sensor information. This data was collected under PX4, with RTK positional data recorded.

Aerial Robotics

Aerial robotics studies the design, perception, control, and coordination of flying robots for navigation, inspection, transport, and physical interaction.

See all 340 papers

Cybflight: An Embedded Rust Autopilot for Aerial Robotics Research
Sep 25, 2026Yifan Lin, Chao Qin, H S Helson Go +1Aerial Robotics
Institute for Aerospace Studies, University of Toronto, Toronto, Canada
Droneulator: A Portable UAV Simulator for Agricultural Workflows with RotorPy and Godot 4
May 22, 2026Jacob Swindell, Michael Lowen, Marija Popovic +1Robotics Simulation Robot Navigation
Lincoln Centre for Autonomous Systems (L-CAS), University of Lincoln, Lincoln, UK. · Faculty of Aerospace Engineering, MAVLab, Delft University of Technology (TU Delft), Delft, Netherlands.
OpenPodcar2: a robust, ROS2 vehicle for self-driving research
Apr 27, 2026Rakshit Soni, Chris Waltham, Md Umar Ibrahim +2Autonomous Driving
School of Engineering and Physical Science, University of Lincoln, UK

Sep 25, 2026J/K move · Enter open · S save

Cybflight: An Embedded Rust Autopilot for Aerial Robotics Research

Yifan Lin, Chao Qin, H S Helson Go, Hugh H. -T. Liu

Institute for Aerospace Studies, University of Toronto, Toronto, Canada

Aerial Robotics

Open paper AskPDF arXiv

Bringing an aerial robotics method from simulation to flight should not require rebuilding a mature autopilot or adding a companion computer. Cybflight is an open-source embedded Rust research autopilot whose typed, replaceable interfaces connect hardware access, perception, state estimation, trajectory planning, and control. This modular development and compile-time optimization workflow is demonstrated with replaceable Rust implementations of model predictive contour-tracking control (MPCTC) and incremental nonlinear dynamic inversion (INDI) running on one STM32H743 without a companion computer. Using this configuration, the vehicle reaches 12.38 m/s during indoor flight, while an outdoor flight using global navigation satellite system (GNSS) position updates reaches 31.4 m/s. These flights show that running demanding estimation and nonlinear control entirely on a flight-controller microcontroller need not come at the expense of a modular autopilot structure.

Abstract

Aerial Robotics

Explore similar work

Cybflight: An Embedded Rust Autopilot for Aerial Robotics Research

Droneulator: A Portable UAV Simulator for Agricultural Workflows with RotorPy and Godot 4

OpenPodcar2: a robust, ROS2 vehicle for self-driving research