← All research storiesEngineering

Peer-reviewedEngineering

A new flight planner keeps drones out of the way of moving obstacles, with a mathematical safety guarantee

MIT engineers built SANDO, a planner that keeps a flying robot inside a constantly updated safe zone, avoiding moving obstacles in simulations and 12 real flights.

How far alongPeer-reviewedReviewed by independent experts for a journal before publication. Peer review is a quality check, not a grade of evidence strength: confidence still depends on study design, limitations, replication and the wider evidence.
How soon could it reach you?A few years outPromising, but needs more testing or engineering first. Our rough editorial guess is several years, with a lot of uncertainty.
Where it comes fromIEEE Transactions on RoboticsMIT; Comillas Pontifical University (Madrid) · Oct 7, 2026Read the original →

What the scientists did

Drones that deliver packages or search disaster zones must dodge things that move, like people, vehicles or other drones, whose future paths are unknown. Most planners either assume obstacles stand still or dodge moving ones without any proof that they’ll stay safe.

The MIT team’s system tracks each moving obstacle and uses its top possible speed to draw a bubble around everywhere it could reach. It then builds a “safety corridor” of connected 3D spaces guaranteed to be free of obstacles, steers away from crowded areas, and picks the fastest path inside the corridor. While flying, it keeps updating the corridor and replanning using only the drone’s own computer and sensors.

Words to know

Trajectory planning

Working out the exact path and speed a robot should follow to reach a goal.

  • A drone choosing how to fly around a tree
  • A robot vacuum mapping its route around furniture

Safety corridor

A chain of connected spaces the robot is allowed to fly through because nothing can be there.

  • Like a tunnel of open air that keeps changing shape as obstacles move

UAV

Uncrewed aerial vehicle, a flying robot with no pilot on board.

  • Quadcopter drones
  • Delivery drones

What they found

In computer simulations, SANDO reached its goal faster than several leading planners and avoided collisions in every test environment. In 12 real test flights with a drone, it avoided all moving obstacles.

12real test flights, all without hitting moving obstacles

Why it matters to you

If drones are going to carry packages over neighborhoods or help firefighters, people need to trust they won’t crash into anyone. Planners with built-in safety guarantees are one step toward drones that can share space with people more safely.

Where this could lead

Safer delivery and inspection drones

Drones could fly through busy spaces while provably keeping clear of moving people and vehicles.

3–10 years

Talk-to-your-drone controls

The team plans to combine the planner with AI models that understand plain-language instructions.

3–10 years

These are possibilities the research points toward, not promises. Most early findings take years of testing before they reach everyday life, and some never do.

Why scientists care

Robotics researchers want guarantees, not just good test results, before robots work around people. SANDO shows a way to get a safety guarantee fast enough to run on a small flying robot in real time. The code is publicly available.

Keep in mind

The guarantee only holds if the drone knows the maximum speed obstacles can reach; something faster than expected could break it. Real-world testing was limited to 12 flights, and the team says the system could be made more computationally efficient. This planner uses math and optimization rather than machine learning.

How far along is the science?

  • Peer-reviewedReviewed by independent experts for a journal before publication. Peer review is a quality check, not a grade of evidence strength: confidence still depends on study design, limitations, replication and the wider evidence.
  • PreprintA manuscript shared publicly, often before or alongside journal review. Check its separate journal and review status (each story says what we know). Results may change.
  • Conference talkPresented to other scientists at a meeting. Usually short and early; a full paper may come later.
  • Agency reportPublished by a science agency such as NASA, NOAA or NIH. Reviewed internally, but not by an outside journal.
  • Press releaseAn announcement from a university or company. Useful, but always check the research it describes.

How soon could it reach you?

  • Could reach you soonCould show up in products, services or advice soon. Our rough editorial guess is about 1 to 3 years, not a forecast from the research.
  • A few years outPromising, but needs more testing or engineering first. Our rough editorial guess is several years, with a lot of uncertainty.
  • Long-term scienceBasic research that builds knowledge. Real-world uses, if any, are far off.
  • Small stepA small update to earlier work. Useful to researchers, but not a change you would notice yet.

Every story links to its original source: the paper itself, or the conference program when no paper is out yet. Explanations are written with AI, usually from the paper’s abstract (each story says what it was written from), then checked against the source. We only use journals, recognized preprint servers, science agencies, scientific conferences and established science news outlets.