New Navigation System Guarantees Collision-Free Drone Flight in Unknown Spaces
Researchers at MIT created a new navigation system called SANDO that helps flying robots avoid moving obstacles in unmapped areas. The system uses math to guarantee that a drone will not crash into anything while finding its destination.
In short: Researchers at MIT created a new navigation system called SANDO that helps flying robots avoid moving obstacles in unmapped areas. The system uses math to guarantee that a drone will not crash into anything while finding its destination.
Imagine a flying robot trying to rescue people deep inside a burning forest or a collapsed building, where tree branches fall and debris shifts without warning. Teaching a flying machine to navigate such a chaotic space without crashing is a major puzzle for engineers, but a new tool from MIT aims to solve it.
What happened, in plain words
Researchers at the Massachusetts Institute of Technology built an autonomous navigation system named SANDO, which stands for Safe AutoNomous trajectory planning for Dynamic unknOwn environments. This tool plans flight paths for uncrewed aerial vehicles, or flying robots. Unlike older systems that only work when an area is fully mapped beforehand or when obstacles stand still, SANDO creates safe travel paths in completely unfamiliar spaces with moving objects. The system uses cameras and sensors on the robot, along with the top speed of nearby obstacles, to calculate safe routes. In computer tests and 12 physical test flights, the system successfully guided drones to their goals without any crashes.
Key points
- Mathematical safety guarantees SANDO is mathematically proven to keep flying robots safe from collisions in completely unmapped environments where obstacles move unpredictably.
- Time-sensitive safety corridors The system builds a series of connected 3D regions that account for the maximum possible speed and future locations of moving obstacles.
- Quick onboard updates Using the robot's onboard computer and sensors, SANDO rapidly recalculates paths as the drone moves so it can react to sudden changes.
- Real-world testing The research team tested the system in 12 flights with a real flying robot, where it successfully avoided all moving obstacles.
Terms explained
- Uncrewed aerial vehicles — Flying machines that carry no human pilot on board and are controlled automatically or from a distance. Example: A small quadcopter flying through a park to take pictures from the sky.
- Trajectory planner — A computer program that figures out the exact path a moving machine should follow to get from one spot to another. Example: A GPS app finding the best driving route from your home to a grocery store.
- Algorithm — A set of step-by-step instructions or rules that a computer follows to solve a problem or complete a task. Example: A recipe that tells you the exact order to mix ingredients to bake a cake.
Why it matters
This technology could eventually help rescue teams explore collapsed buildings, search through hidden tunnels in mines, or deliver packages safely across busy neighborhoods.
What we still don't know
The research is at an early stage, and future work is still needed to make the system more computationally efficient and to combine it with machine-learning models.
Based on reporting from MIT News. This is an independent explainer, written in our own words with AI assistance; MIT News has not reviewed or endorsed it. Read the original for the full details.