# Curriculum-based Sample Efficient Reinforcement Learning for Robust Stabilization of a Quadrotor

arXiv: https://arxiv.org/abs/2501.18490 · HTML: https://mlsanity.com/paper/2501.18490/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in RL for robot navigation and UAV control, one of 15 directions in Deep reinforcement learning methods.
- Builds on 18 of its 36 references that are in the map.
- First posted on arXiv 30 Jan 2025.

## Placed in

- Deep reinforcement learning methods / [RL for robot navigation and UAV control](https://mlsanity.com/deeprl/rl-for-robot-navigation-and-uav-control/)

## Builds on (most cited first, 18 of 18)

- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [Continuous control with deep reinforcement learning](https://mlsanity.com/paper/1509.02971/) (2015)
- [Trust Region Policy Optimization](https://mlsanity.com/paper/1502.05477/) (2015)
- [Curriculum Learning for Reinforcement Learning Domains: A Framework and Survey](https://mlsanity.com/paper/2003.04960/) (2020)
- [Control of a Quadrotor with Reinforcement Learning](https://mlsanity.com/paper/1707.05110/) (2017)
- [Reinforcement Learning for UAV Attitude Control](https://mlsanity.com/paper/1804.04154/) (2018)
- [A Benchmark Comparison of Learned Control Policies for Agile Quadrotor Flight](https://mlsanity.com/paper/2202.10796/) (2022)
- [OmniDrones: An Efficient and Flexible Platform for Reinforcement Learning in Drone Control](https://mlsanity.com/paper/2309.12825/) (2023)
- [Learning to Fly in Seconds](https://mlsanity.com/paper/2311.13081/) (2023)
- [A Learning-based Quadcopter Controller with Extreme Adaptation](https://mlsanity.com/paper/2409.12949/) (2024)
- [End-to-end Neural Network Based Quadcopter control](https://mlsanity.com/paper/2304.13460/) (2023)
- [Aerial Gym -- Isaac Gym Simulator for Aerial Robots](https://mlsanity.com/paper/2305.16510/) (2023)
- [Learning Stabilization Control from Observations by Learning Lyapunov-like Proxy Models](https://mlsanity.com/paper/2303.02215/) (2023)
- [Survey of Simulators for Aerial Robots: An Overview and In-Depth Systematic Comparisons](https://mlsanity.com/paper/2311.02296/) (2023)
- [Design of a Trajectory Tracking Controller for a Nanoquadcopter](https://mlsanity.com/paper/1608.05786/) (2016)
- [The Power of Input: Benchmarking Zero-Shot Sim-To-Real Transfer of Reinforcement Learning Control Policies for Quadrotor Control](https://mlsanity.com/paper/2410.07686/) (2024)
- [Using Simulation Optimization to Improve Zero-shot Policy Transfer of Quadrotors](https://mlsanity.com/paper/2201.01369/) (2022)
- [Towards Real-World Efficiency: Domain Randomization in Reinforcement Learning for Pre-Capture of Free-Floating Moving Targets by Autonomous Robots](https://mlsanity.com/paper/2406.06460/) (2024)

## Built on by (most cited first, 1 of 1)

- [AbideGym: Turning Static RL Worlds into Adaptive Challenges](https://mlsanity.com/paper/2509.21234/) (2025)

## Abstract (arXiv)

This article introduces a novel sample-efficient curriculum learning (CL) approach for training an end-to-end reinforcement learning (RL) policy for robust stabilization of a Quadrotor. The learning objective is to simultaneously stabilize position and yaw-orientation from random initial conditions through direct control over motor RPMs (end-to-end), while adhering to pre-specified transient and steady-state specifications. This objective, relevant in aerial inspection applications, is challenging for conventional one-stage end-to-end RL, which requires substantial computational resources and lengthy training times. To address this challenge, this article draws inspiration from human-inspired curriculum learning and decomposes the learning objective into a three-stage curriculum that incrementally increases task complexity, while transferring knowledge from one stage to the next. In the proposed curriculum, the policy sequentially learns hovering, the coupling between translational and rotational degrees of freedom, and robustness to random non-zero initial velocities, utilizing a custom reward function and episode truncation conditions. The results demonstrate that the proposed CL approach achieves superior performance compared to a policy trained conventionally in one stage, with the same reward function and hyperparameters, while significantly reducing computational resource needs (samples) and convergence time. The CL-trained policy's performance and robustness are thoroughly validated in a simulation engine (Gym-PyBullet-Drones), under random initial conditions, and in an inspection pose-tracking scenario. A video presenting our results is available at https://youtu.be/9wv6T4eezAU.