# Learning to Fly via Deep Model-Based Reinforcement Learning

arXiv: https://arxiv.org/abs/2003.08876 · HTML: https://mlsanity.com/paper/2003.08876/

## Key facts

- Cited by 20 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 19 Mar 2020.

## Built on by (most cited first, 20 of 20)

- [DayDreamer: World Models for Physical Robot Learning](https://mlsanity.com/paper/2206.14176/) (2022)
- [Consciousness in Artificial Intelligence: Insights from the Science of Consciousness](https://mlsanity.com/paper/2308.08708/) (2023)
- [Learning to Fly in Seconds](https://mlsanity.com/paper/2311.13081/) (2023)
- [Dream to Fly: Model-Based Reinforcement Learning for Vision-Based Drone Flight](https://mlsanity.com/paper/2501.14377/) (2025)
- [Mind the Gap when Conditioning Amortised Inference in Sequential Latent-Variable Models](https://mlsanity.com/paper/2101.07046/) (2021)
- [Accelerating Model-Based Reinforcement Learning with State-Space World Models](https://mlsanity.com/paper/2502.20168/) (2025)
- [Action Inference by Maximising Evidence: Zero-Shot Imitation from Observation with World Models](https://mlsanity.com/paper/2312.02019/) (2023)
- [Latent Action World Models for Control with Unlabeled Trajectories](https://mlsanity.com/paper/2512.10016/) (2025)
- [Tracking and Planning with Spatial World Models](https://mlsanity.com/paper/2201.10335/) (2022)
- [Constrained Latent Action Policies for Model-Based Offline Reinforcement Learning](https://mlsanity.com/paper/2411.04562/) (2024)
- [Training microrobots to swim by a large language model](https://mlsanity.com/paper/2402.00044/) (2024)
- [Latent Matters: Learning Deep State-Space Models](https://mlsanity.com/paper/2602.23050/) (2026)
- [Filter-Aware Model-Predictive Control](https://mlsanity.com/paper/2304.10246/) (2023)
- [Deep reinforcement learning for tracking a moving target in jellyfish-like swimming](https://mlsanity.com/paper/2409.08815/) (2024)
- [Multi-Robot Collaboration through Reinforcement Learning and Abstract Simulation](https://mlsanity.com/paper/2503.05092/) (2025)
- [Designing a Robust Low-Level Agnostic Controller for a Quadrotor with Actor-Critic Reinforcement Learning](https://mlsanity.com/paper/2210.02964/) (2022)
- [PRISM: Probabilistic Real-Time Inference in Spatial World Models](https://mlsanity.com/paper/2212.02988/) (2022)
- [Learning Representative Trajectories of Dynamical Systems via Domain-Adaptive Imitation](https://mlsanity.com/paper/2304.10260/) (2023)
- [Model-Free versus Model-Based Reinforcement Learning for Fixed-Wing UAV Attitude Control Under Varying Wind Conditions](https://mlsanity.com/paper/2409.17896/) (2024)
- [WOLF: World Model Guided LiDAR Exploration with Predictive Frontiers](https://mlsanity.com/paper/2609.23656/) (2026)

## Abstract (arXiv)

Learning to control robots without requiring engineered models has been a long-term goal, promising diverse and novel applications. Yet, reinforcement learning has only achieved limited impact on real-time robot control due to its high demand of real-world interactions. In this work, by leveraging a learnt probabilistic model of drone dynamics, we learn a thrust-attitude controller for a quadrotor through model-based reinforcement learning. No prior knowledge of the flight dynamics is assumed; instead, a sequential latent variable model, used generatively and as an online filter, is learnt from raw sensory input. The controller and value function are optimised entirely by propagating stochastic analytic gradients through generated latent trajectories. We show that "learning to fly" can be achieved with less than 30 minutes of experience with a single drone, and can be deployed solely using onboard computational resources and sensors, on a self-built drone.