STARMAC Quadrotor Helicopter UAV Project

STARMAC, the Stanford Testbed of Autonomous Rotorcraft for Multi-Agent is a testbed of 6 quadrotor helicopter unmanned aerial vehicles (UAV) that fly indoors and outdoors to experiment with autonomous agent algorithms. This video presents an overview of the control system, the vehicle capabilities, and the experimental applications for the testbed. Results of many experimental flights are shown, along with some fun flight results that push the limits of what the vehicles should be able to handle.

Category:

Science & Technology

Tags:

• autonomous
• control
• distributed
• flight
• helicopter
• multi-agent
• path
• planning
• quadrotor
• rotorcraft
• starmac
• system
• tests
• uav
• aviation
• space
• electronics
• computer

[zephyr9673]

Could you use Bluetooth or other wi fi to determine distances between swarm members?

[gabehoffmann]

@zephyr9673 Yes, that's a possibility. It's a decent challenge to implement such a system right now though, but there have been several research papers that demonstrated this capability. Relative positioning is much easier with specialized radios or GPS.

[DrJez1]

Amazing. May I ask how you were involved in this project?

[gabehoffmann]

Thanks for the kind words. It was a part of the work for my PhD, working toward programming vehicles to understand how to use their sensing capabilities to achieve their goals. Personally, I did a bit of everything on this; design, fabrication, theory, programming, system id, experiments. Steve Waslander and I had several masters students helping us, and several PhD students came up to speed on it, contributed substantially, and carried it on now that we graduated.

[jasanpahaf]

for how long can you keep it in air? and about the GPS... does it uses way more energy than with out it?

what's its range?

[gabehoffmann]

@jasanpahaf when it's configured as shown, about 15-20 minutes. The embedded GPS receiver barely consumes any power relative to the rest of the electronics, so it doesn't use way more energy with GPS. When you fill up the payload with batteries, the predicted flight time is on the order of 1 hour, though we haven't tried that. A safe flight speed is 25 mph, so that puts the max range at somewhere around 25 miles, with a decent amount of uncertainty.

[gabehoffmann]

The forces from blade flapping increase with flight speed. At hover it has essentially no effect. It becomes more and more noticeable as speed increases. Momentum theory equations were accurate for predicting steady state thrust and still approximately correct for some transient conditions, but not all transient conditions. Other effects can be quite important, such as where the vortices go when they are shed from the propellors.

[r55741]

*cocks shotgun*

them birds aint coming in my yard!

[skillguitar]

@steam4o True.. But I meant a Hoax .. :)

[Lantho650]

Im building A PROJECT A MAKING A ufo hoax

[jjppmm29]

Ahhh yes hover drone Ballet 3:30

[skillguitar]

Make it bigger and u have a UFO...

[DrJez1]

@gabehoffmann I'm in my fourth year of mech. eng. about to start a 9 months project with 3 other people. Constructing a quadrotor is a one option (out of 80!). It would include all the control theory and programming. Would you say most time was spent typing code? How many people were involved? Thanks, Jez