Aero-VPA for drones

The VPA architecture trained on flight data: the drone decides from its own camera, with no map, and a safety shield remembers the obstacles it has already passed.

Status: Research prototype tested in simulated flights (PX4 SITL with Gazebo Harmonic). Code, trained weights and every benchmark record are public; flights on a real drone are planned for the first half of 2027.

Top view of a forest: the drone's camera view, the path flown so far, two remembered trees and the remaining route to the goal 1 2 3 Camera view 15 s memory Start Goal
Seen from above: the camera's view, the path flown so far and trees kept in memory for 15 s. The map is only for you; the drone flies from its camera alone.

How it works

1. World model

A Vision Transformer embeds each RGB-D camera frame (160 × 128). A JEPA predictor, trained to predict the next embedding rather than pixels, forecasts the scene 0.1 s ahead; the disagreement of its five heads measures uncertainty.

2. Actor

A primitive selector and a conditional flow-matching network turn the latents, the next waypoint and the vehicle state into a 16-step velocity and yaw-rate chunk (0.32 s) in two integration steps. A decision is five network calls, whatever the chunk length.

3. Safety shield

A deterministic runtime filter keeps 15 s of depth returns in a local 3-D memory, so trees beside and behind the 73° camera stay known. It limits the approach speed to every remembered obstacle, caps motion outside the field of view and backs off along the flown path when too close. It uses no map.

The world model and the actor are the same Vision-Predictor-Action design as VPA for robot arms: no language decoder in the control loop, so the cost of a decision does not grow with the length of the motion.

Results in simulation

Goal reached, recommended policy (r1)
52.2% with the shield, 21.1% without94 of 180 missions, 95% CI 45.0–59.4%; paired exact McNemar p = 1.2 × 10⁻¹⁰.
Goal reached, all six policies
42.8–52.2% with the shieldEvery checkpoint improved significantly with the shield.
Crashes
From 72–82% down to 22–43%Across the six checkpoints; r1 goes from 72% to 22%, i.e. 78% of its flights end without a crash.
Test setup
36 test worlds × 5 missionsForest worlds of three densities, disjoint from training; every policy flies the same missions.
Model size
18.13 million parametersAbout 36 MB at FP16 (72 MB at FP32).
Control rate
100 Hz velocity setpoints to PX4Asynchronous two-rate stack designed for a 15 W NVIDIA Jetson Orin Nano.

PX4 SITL with Gazebo Harmonic. The model card on Hugging Face has the per-checkpoint and per-density tables, and sim/paired_stats.py in the code repository recomputes every number from the published records. These are simulation results with one training seed per recipe.

Parameter budget

Module Parameters Role
Vision encoder 11.17M 4-channel RGB-D (128 × 160) Vision Transformer to a 256-dim latent
Primitive selector 0.67M Chooses one of 6 flight primitives
JEPA predictor 2.72M 5-head ensemble forecasting the next latent and its uncertainty
Flow-matching solver 3.57M 16-knot, 4-DoF heading-frame velocity chunk
Total 18.13M About 36.3 MB of FP16 weights

How customers use it

On-site software

Customers
Factories, logistics centers and drone operators
Price
20% of the labor cost the software saves
Sales path
Paid proof of concept (PoC), then an annual subscription

Technology licensing

Customers
Robot, drone and chip makers
Price
Royalty per unit shipped
Sales path
The free open release builds adoption; products sold with VPA inside need a paid license

Open release and licensing

For commercial integration, proprietary licensing or defense adaptation, contact SLING AI.

Contact us about licensing