Scientists Taught a Toy Drone to Fly With Brain-Like Software

On October 1, 2026, the Aerial Robotics Working Group — a community of drone and robotics fans — is holding a free, one-hour online meeting. The featured talk comes from Anel Olivares Cruz, a researcher at the Autonomous University of the State of Hidalgo in Mexico. She'll explain how she built three different flight controllers that steer a drone using software inspired by the human brain, rather than the conventional equations most drones rely on today.

The brain-like part is called a 'spiking neural network' — think of it as code where signals fire in short pulses, similar to how real neurons communicate. The appeal is efficiency: these systems can make fast decisions while using much less computing power, which matters on a drone where every gram of battery and every chip adds weight and cost. She built the controllers using Nengo, a free tool for prototyping brain-style software.

To test them, she used a DJI Tello — a palm-sized drone that costs around $100 at retail. The drone was given simple flying tasks: follow a set path, stay on course, and correct itself in real time. That detail matters. This wasn't a simulation or a lab-only demo; it was a physical drone you or I could buy, reacting live. Her code and research profile are posted publicly for anyone curious to dig in.

Here's the honest caveat: this is an early-stage academic talk, not a product launch. A toy drone following a path indoors is a long way from a delivery drone navigating a windy city. The real value is the direction of travel — if brain-inspired control works, future drones could fly longer, need cheaper hardware, and handle surprises better. That eventually touches package delivery, farm spraying, search-and-rescue, and inspection of bridges or power lines. If any of that interests you, the meeting is free and open to anyone with the link.