AI-Based Obstacle Avoidance for UAV Path Planning
Modern UAV path planning systems process up to 30 million distance measurements per second while maintaining obstacle detection rates under 50 milliseconds. These systems must balance computational constraints with environmental complexity—from structured indoor environments with defined obstacles to dynamic outdoor scenarios with unpredictable moving objects and varying weather conditions that affect sensor reliability.
The fundamental challenge lies in balancing computational efficiency against the robustness required for real-time decision-making in unpredictable environments.
This page brings together solutions from recent research—including sensor fusion architectures that integrate lidar and camera data, hierarchical planning systems that separate global routing from local collision avoidance, dynamic map updating mechanisms for tracking moving obstacles, and reliability-based path selection that considers population density in urban environments. These and other approaches offer practical implementations for UAVs operating across diverse mission profiles while maintaining safety and efficiency standards.
1. Autonomous Drone with Sensor Fusion for 3D Mapping and Obstacle Avoidance
KOREA ADVANCED INSTITUTE OF SCIENCE AND TECHNOLOGY, 2025
Autonomous drone system for exploration and reconnaissance in unknown environments. The system allows the drone to autonomously fly, map the environment, locate targets, avoid obstacles, and return home. The drone acquires data from cameras, lidar, and IMU to estimate pose, recognize targets, and generate a 3D map. It plans safe paths using ray-casting and sensor fusion. The drone applies the path to fly autonomously. This allows it to explore unknown areas, accurately locate targets, avoid obstacles, and return home.
2. Hierarchical Multi-Drone and Sensor Platform with Federated Path Planning and Information Lateralization
SOTER TECHNOLOGY INC, 2024
Facilitating managing of paths for unmanned vehicles using a hierarchical multi-drone/sensor platform with information lateralization and federated path planning. The platform involves multiple drones, ground robots, and sensors with complementary functions acquiring heterogeneous information at different resolutions. This information is integrated and fed back to adjust planned paths for missions. The platform is modeled after brain lateralization, where drones/sensors have specialized roles like human hemispheres.
3. Autonomous Mobile Robot Control System with Real-Time Route Adjustment Using Environmental Change Detection and Path Prediction
TOYOTA JIDOSHA KABUSHIKI KAISHA, 2024
An autonomous mobile robot control system that updates its route plan in real-time based on environmental changes detected by cameras. The system continuously monitors its surroundings, detects moving objects, predicts their paths, and generates avoidance procedures to prevent collisions. This enables the robot to dynamically adjust its route to avoid obstacles and minimize disruptions to human traffic.
4. 3D Navigation Method Using Predefined Spatial Segments with Unique Coordinate Combinations
AIRBUS DEFENCE AND SPACE GMBH, 2024
A navigation method for objects in 3D space that reduces computational requirements by representing the environment as a set of predefined spatial segments, each defined by a volume and a unique coordinate combination, and determining object positions and routes within these segments.
5. Autonomous Drone Navigation System with Dynamic Object Tracking and Map Updating for Indoor Path Planning
ALARM.COM INC, 2024
Improving autonomous navigation of drones inside buildings by dynamically tracking objects and their status to generate accurate maps for path planning. The drone obtains a map of the building with locations of dynamic objects. It periodically updates the map with current object status to capture changes. When instructed to perform an action, the drone computes the route based on the dynamic map and object status to avoid obstructions. This allows the drone to adapt and reroute if objects move or block paths.
6. UAV Flight Planning System Utilizing Population Density and Ground Condition-Based Reliability Mapping
AMAZON TECHNOLOGIES INC, 2024
System for safe UAV flight planning over populated areas that uses localized population density and ground condition data to select routes that avoid densely populated areas. The system generates a reliability map of a region by dividing it into cells and scoring each cell based on population density and ground conditions. A search algorithm uses the reliability map to plan UAV routes that minimize overflight of populated areas.
7. Marine Navigation System with AI-Based Collision Risk Evaluation and Route Adjustment Mechanism
FURUNO ELECTRIC CO, 2023
A marine navigation system for safe ship navigation using artificial intelligence. The system includes a navigation route planning apparatus that receives planned routes, vessel and obstacle information, and uses machine learning to evaluate collision risks and generate avoidance routes. The apparatus determines whether to deviate from the planned route based on the collision risk assessment, enabling safe navigation through dynamic ocean environments.
8. Route Optimization Method for Mobile Robots Utilizing A* Global Planning and Dynamic Window Local Optimization
STAR INSTITUTE OF INTELLIGENT SYSTEMS, CHONGQING UNIVERSITY, DIBI INTELLIGENT TECHNOLOGY RESEARCH INSTITUTE CO LTD, 2023
A route optimization method for mobile robots that combines global planning using A* algorithm with local optimization using dynamic window algorithm. The method first generates a global route using A*, then filters out redundant nodes to create a key node-based route. Local optimization is then performed on each segment of the route using dynamic window algorithm to improve route smoothness and safety.
9. Electric Aircraft Flight Controller with Autonomous Collision Avoidance Using Sensor-Driven Navigation Adjustments
BETA AIR LLC, 2023
Automated sense and avoid system for electric aircraft, comprising a flight controller that receives sensor inputs, identifies potential collision threats, generates navigation status, produces flight modifications, and autonomously initiates those modifications to prevent collisions.
10. Cell-Based Ground Obstruction Mapping for Drone Route Planning Using Dynamic Occupancy Metrics
AMAZON TECHNOLOGIES INC, 2023
Planning efficient and safe routes for drones based on localized ground obstruction data. The method involves dividing a region into cells and calculating occupancy metrics for each cell based on data like building footprints. Cells with high occupancy metrics are labeled as obstructed. The cells are further divided until they can be labeled obstructed or free. This dynamic and granular map of ground obstructions is used to plan drone paths through the free cells.
11. Flight Device with Dynamic Course Adjustment for Obstacle Density Avoidance
KDDI CORP, 2023
Flight device that dynamically adjusts its flight plan to avoid high-density areas of obstacles, such as cities or crowds, by proactively changing course when approaching areas with a threshold density of obstacles. The device continuously monitors its surroundings and adjusts its flight path to maintain a safe distance from obstacles, even when flying over densely populated areas.
12. Method for Generating Pathing Data Using Real-Time Sensor Input Over High-Throughput Network
HAND HELD PRODUCTS INC, 2023
A method for generating optimized pathing data for traveling objects in an environment using continuous real-time sensor data transmitted over a high-throughput communications network. The method receives real-time sensor data, determines the environment's status based on the data, and generates optimized pathing data for one or more traveling objects based on the status. The optimized pathing data is then transmitted to the traveling objects' control devices to enable autonomous navigation.
13. System for UAV Delivery Scheduling and Routing Utilizing Multi-Source Travel Data Integration
AMAZON TECHNOLOGIES INC, 2023
Optimizing scheduling and routing of deliveries by unmanned aerial vehicles (UAVs) using travel-related data from UAV sensors, data aggregators, weather services, and obstacle databases. The system receives data on obstacles, weather, crowds, interference, etc. from UAV sensors and other sources to evaluate and update flight plans. The UAV data is combined with other sources and stored for analysis.
14. Vehicle Motion Planning System with Nominal and Abort Trajectory Determination Based on Real-Time Movement and Surrounding State Analysis
APTIV TECHNOLOGIES LTD, 2023
Planning a motion of a vehicle that minimizes the risk of accident. The method includes determining a nominal trajectory for the vehicle based on a desired maneuver to be carried out in a traffic space, on a current state of movement of the vehicle and on a detected state of a surrounding of the vehicle, and determining, via the motion planning module of the control system, an abort trajectory branching off from the nominal trajectory and guiding the vehicle to a safe condition regardless of the desired maneuver.
15. Drone System with Stereo Vision and Depth Sensors for Autonomous Navigation and Obstacle Avoidance
DIGIT7 INDIA PRIVATE LTD, 2023
A drone system for inventory management in warehouses, enabling autonomous navigation and obstacle avoidance through a combination of stereo vision, optical flow, and depth sensors. The system generates a 3D map of the environment, estimates the drone's spatial position and orientation, and uses a collision prevention feature to find the shortest path between nodes while avoiding static and dynamic obstacles.
16. Autonomous Aerial Vehicle Navigation Using Crowd-Sourced Dynamic Obstacle Data with Airspace Partitioning
HERE GLOBAL BV, 2023
Enabling autonomous aerial vehicles to safely navigate complex environments by leveraging crowd-sourced data on dynamic obstacles they encounter during their flights. The system partitions the airspace into 3D shapes and records when aerial vehicles enter and exit each shape to build up a database of dynamic obstacle movements. This data is processed to compute collision probabilities for each shape. Aerial vehicles can then request routes that avoid shapes with high collision probabilities.
17. Route Planning Method for Mobile Robots Utilizing A* Global Pathfinding and Dynamic Window Local Optimization
DIBI INTELLIGENT TECHNOLOGY RESEARCH INSTITUTE CO LTD, 2022
A route optimization method for mobile robots that combines global planning with local optimization. The method first uses A* algorithm to plan a global route, then filters out redundant nodes to retain key nodes. Local optimization is then performed using the dynamic window algorithm to improve route smoothness and safety. This hybrid approach enables efficient and safe robot movement by leveraging the strengths of both global and local planning methods.
18. Distributed Architecture for Safety-Level Integration in Unmanned Aerial Vehicles with Non-Deterministic Algorithms
KUTTA TECHNOLOGIES INC, 2022
Enabling safe, autonomous flight of unmanned aerial vehicles (UAVs) using a distributed architecture that allows non-deterministic AI/ML algorithms on UAVs while maintaining safety. The architecture involves multiple safety levels assigned to individual functions. Vehicles with lower safety levels can connect to vehicles with higher levels to operate at the higher level. This promotes safety by leveraging trusted systems. It also allows cost-reduced, low-cost UAVs to operate in restricted environments by connecting to higher safety level vehicles.
19. Aerial Robotic Navigation System with Trajectory Prediction and Collision Detection Using Object Tracking and Real-Time Route Management
EVERSEEN LTD, 2022
A navigation system for aerial robotic devices that enables obstacle avoidance through trajectory prediction and collision detection. The system employs a combination of object tracking, prediction algorithms, and real-time route management to predict the trajectory of moving objects. It continuously updates the prediction list by filtering out measurement noise and determining the velocity and acceleration vectors of each object. The system then generates predicted trajectory points for each object in the prediction list, enabling the robotic device to avoid obstacles by predicting their future positions. The system maintains a tracking list of detected objects with their corresponding trajectory points, allowing the robotic device to maintain a continuous view of the environment and avoid collisions.
20. Aerial Vehicle Routing Method Using 3D Grid-Based Shape Representation with Probability-Based Collision Optimization
HERE GLOBAL BV, 2022
Method for routing an aerial vehicle through a 3D space by representing the space as a grid of three-dimensional shapes, matching the vehicle's start and target locations to corresponding shapes, and computing a collision-optimized route based on probability data for each shape. The route is evaluated based on collision probability, enabling autonomous aerial vehicles to safely navigate complex environments.
Get Full Report
Access our comprehensive collection of 45 documents related to this technology