Advanced Pathfinding Strategies for Drone Navigation

Researchers at Hong Kong University MaRS Lab have just published another jaw dropping paper featuring their safety-assured high-speed aerial robot path planning system dubbed "SUPER". With a single MID360 lidar sensor they repeatedly achieved autonomous one-shot navigation at speeds exceeding 20m/s in obstacle rich environments. Since it only requires a single lidar these vehicles can be built with a small footprint and navigate completely independent of light, GPS and radio link. This is not just #SLAM on a #drone, in fact the SUPER system continuously computes two trajectories in each re-planning cycle—a high-speed exploratory trajectory and a conservative backup trajectory. The exploratory trajectory is designed to maximize speed by considering both known free spaces and unknown areas, allowing the drone to fly aggressively and efficiently toward its goal. In contrast, the backup trajectory is entirely confined within the known free spaces identified by the point-cloud map, ensuring that if unforeseen obstacles are encountered or if the system’s perception becomes uncertain, the system can safely switch to a precomputed, collision-free path. The direct use of LIDAR point clouds for mapping eliminates the need for time-consuming occupancy grid updates and complex data fusion algorithms. Combined with an efficient dual-trajectory planning framework, this leads to significant reductions in computation time—often an order of magnitude faster than comparable SLAM-based systems—allowing the MAV to operate at higher speeds without sacrificing safety. This two-pronged planning strategy is particularly innovative because it directly addresses the classic speed-safety trade-off in autonomous navigation. By planning an exploratory trajectory that pushes the speed envelope and a backup trajectory that guarantees safety, SUPER can achieve high-speed flight (demonstrated speeds exceeding 20 meters per second) without compromising on collision avoidance. If you've been tracking the progress of autonomy in aerial robotics and matching it to the winning strategies emerging in Ukraine, it's clear we're likely to experience another ChatGPT moment in this domain, very soon. #LiDAR scanners will continue to get smaller and cheaper, solid state VSCEL based sensors are rapidly improving and it is conceivable that vehicles with this capability can be built and deployed with a bill of materials below $1000. Link to the paper in the comments below.

UAV Path Planning Based on Deep Reinforcement Learning This approach enables drones to navigate complex environments autonomously and efficiently. DRL-based systems dynamically learn optimal paths, balancing safety, speed, and energy use for various applications. How It Works? UAVs equipped with DRL algorithms explore environments, learning to avoid obstacles, optimize flight paths, and adapt to dynamic conditions. The system uses convolutional neural networks (CNNs) for spatial awareness, reinforcement learning frameworks like PPO, DDPG, or SAC for decision-making, and reward functions to guide actions, penalizing collisions and rewarding efficient navigation. Recurrent networks manage sequential dependencies for real-time adaptability. Applications - Disaster response: Efficient delivery of supplies in challenging terrains. - Surveillance: Monitoring large areas with minimal human intervention. - Agriculture: Precision spraying and crop health monitoring. Key Challenges and Solutions - Dynamic Obstacles: Solved with predictive models and real-time path adjustments. - Computation: Addressed by deploying DRL models on edge hardware with optimized frameworks. - Generalization: Improved through extensive training in diverse simulated environments. Implementation Steps - Environment Setup: Design virtual environments replicating real-world scenarios. - Data Preparation: Use synthetic and real flight data to simulate varying conditions. - Training: Employ DRL algorithms like PPO or DDPG with custom reward functions for navigation tasks. - Simulation Testing: Validate policies in simulation environments with diverse obstacles. - Deployment: Optimize DRL models for UAV onboard processors # Idea of implementation Environment Setup: - Create a simulation with UAV dynamics (position, velocity) and obstacles. - - Define observation space (sensor data like LiDAR, camera frames) and action space (control commands such as thrust, pitch, roll). Reward Function: - Reward the UAV for moving toward the goal and penalize it for collisions, straying from the path, or idling. The reward function combines these factors to guide the agent toward optimal behavior. In example: R=α⋅Δdistance−β⋅collision penalty−γ⋅time penalty Training the DRL Agent: - Use algorithms like Proximal Policy Optimization (PPO) or Deep Deterministic Policy Gradient (DDPG). The UAV learns by interacting with the environment, selecting actions, and updating the policy based on rewards received after each action. Simulation Testing: - After training, test the agent’s performance in various scenarios, measuring goal completion time, collision rates, and efficiency. Fine-tune if necessary. Deployment: - Optimize the trained model for edge devices (e.g., NVIDIA Jetson) and test it on real UAV hardware. Implement safety protocols and real-world testing to ensure safe operation. Follow me for insights on AI, ML, and Robotics

Nature just published our article on Bee-Nav, a strikingly efficient robot navigation strategy inspired by honeybee learning flights 🐝->🚁. The strategy enables even small drones to travel for hundreds of meters and then successfully return home, while using only a tiny 42-kB neural network 🧠. A link to the open access article is included in the comments below. Most autonomously navigating robots need extensive computing 💻 to construct highly detailed 3D maps 🗺. In contrast, #honeybees 🐝 have been solving the problem of navigation with tiny brains for already millions of years. In Bee-Nav we leverage biological knowledge about honeybee navigation, while filling in some of the blanks. Specifically, just like honeybees, our robot first performs a short learning flight near its home. During this flight, the robot gathers panoramic images of its environment. It then trains a tiny neural network to form a #view #memory that maps the images to the direction and distance home: 🖼️-->🧠-->🧭,📏. The targets for learning come from the robot’s own, noisy #path #integration: keeping track of where it is by integrating its movement speed and direction over time 🐝 💨🧭. After the learning flight, the robot can immediately fly far away to perform its tasks, while doing path integration. When it decides to return home, it will come back in a straight line to the supposed home position. Upon arrival, there will be an offset to the home location due to path integration drift 🚁<--->🏠. However, as long as the robot ends up in the learned homing area, it can use its neural network to cancel the drift and come home 🎯. We show with simulation experiments that the learned homing area can be very small compared to the total flight area °↔⚪ (~4% without and ~0.25% with a compass 🧭). Moreover, using path integration for learning the view memory is no problem for coming home. However, it does lead to winding paths within the learned homing area 〰️🏡. In robotic experiments, Bee-Nav enabled a small drone to navigate over hundreds of meters in various environments with tiny neural networks. Congratulations to the first author, Dequan Ou, with his first scientific article 🤯👏  and all other co-authors Jesse H., Maciek Jankowski, Michiel Firlefyn, Christophe De Wagter (all at TU Delft | Aerospace Engineering), Florian Muijres (from Wageningen University & Research), and Jacqueline Degen (from Carl von Ossietzky University of Oldenburg) – it was an exhilarating ride 😅 I think it is worth emphasizing that much of the work for the article has been done in MSc thesis projects 🎓, while another larger part has been funded by NWO (Dutch Research Council) in the context of my NWO VICI grant on neuromorphic learning for advanced insect-inspired AI (NL-AI²). 

Built an autonomous drone mission planner that updates flight paths mid-air. Tested it on ArduPilot SITL and it actually worked. The goal was simple: program a drone to fly 15 waypoints, then, after the 10th one, make it fly perpendicular for 100m before continuing. All while the drone is already in the air. Getting DroneKit to talk to ArduPilot #SITL (Software In The Loop) was straightforward. The tricky part was calculating that perpendicular waypoint. Had to use Haversine formulas for distance and bearing, then inject it into an active mission via #MAVLink commands without the drone landing. Built this in Python with: - DroneKit for flight control - MAVLink protocol for waypoint commands - Real-time telemetry logging (distance, ETA, current position) - Dynamic mission updates mid-flight Also implemented a 3D path planning system for multiple drones. If 3 drones need to fly different routes, they can't be at the same point at the same time. Traditional A* doesn't handle this. Modified A*, where each grid point tracks occupation times. If drone A is at point (50,50,50) at t=10s, drone B either routes around it or delays. Tested on a 100×100×100 weighted grid with 3 simultaneous paths. Everything runs in simulation. ArduPilot SITL handles the #physics, #MAVProxy manages connections. Code is on GitHub with setup instructions for both systems. Links in comments. #Drones #Robotics #PathPlanning

QUAV: Quantum-Assisted Path Planning and Optimization for UAV Navigation with Obstacle Avoidance by New York University Abu Dhabi & Thales by Nouhaila I., Muhammad Kashif, Alberto Marchisio, Yung-Sze Gan, Frédéric Barbaresco , Muhammad Shafique https://lnkd.in/eiQZUnBN Abstract The growing demand for drone navigation in urban and restricted airspaces requires real-time path planning that is both safe and scalable. Classical methods often struggle with the computational load of high-dimensional optimization under dynamic constraints like obstacle avoidance and no-fly zones. This work introduces QUAV, a quantum-assisted UAV path planning framework based on the Quantum Approximate Optimization Algorithm (QAOA), to the best of our knowledge, this is one of the first applications of QAOA for drone trajectory optimization. QUAV models pathfinding as a quantum optimization problem, allowing efficient exploration of multiple paths while incorporating obstacle constraints and geospatial accuracy through UTM coordinate transformation. A theoretical analysis shows that QUAV achieves linear scaling in circuit depth relative to the number of edges, under fixed optimization settings. Extensive simulations and a realhardware implementation on IBM’s ibm_kyiv backend validate its performance and robustness under noise. Despite hardware constraints, results demonstrate that QUAV generates feasible, efficient trajectories, highlighting the promise of quantum approaches for future drone navigation systems.

Read about our work on "A Tree-based Next-best-trajectory Method for 3D UAV Exploration" published at the IEEE Transactions on Robotics from Björn Lindqvist, Akash Patel, Kalle Löfgren. Link: https://lnkd.in/dhA3EX99 This work presents a fully integrated tree-based combined exploration-planning algorithm: Exploration-RRT (ERRT). The algorithm is focused on providing real-time solutions for local exploration in a fully unknown and unstructured environment while directly incorporating exploratory behavior, robot-safe path planning, and robot actuation into the central problem. ERRT provides a complete sampling and tree-based solution for evaluating “where to go next” by considering a trade-off between maximizing information gain, and minimizing the distances travelled and the robot actuation along the path. The complete scheme is evaluated in extensive simulations, comparisons, as well as real-world field experiments in constrained and narrow subterranean and GPS-denied environments. The framework is fully ROS-integrated, straight-forward to use, and we open-source it at https://lnkd.in/dRruwyYh. #robotics #AI #autonomy #exploration