Slashdot Mirror
MIT Drone Autonomously Avoids Obstacles At 30 MPH (roboticstrends.com)
An anonymous reader writes: Traditional obstacle-avoidance software uses images from each camera, and search through the depth-field at multiple distances to determine if an object is in the drone's path. Such approaches are computationally intensive, meaning the drone can't fly faster than 6 miles per hour without specialized processors. PhD student at MIT’s Computer Science and Artificial Intelligence Lab, Andrew Barry realized that at the speeds his drone could travel, the world simply does not change much between frames. Because of that, he could get away with computing just a small subset of measurements — distances of 10 meters away. "As you fly, you push that 10-meter horizon forward, and, as long as your first 10 meters are clear, you can build a full map of the world around you," Barry says.
Itâ(TM)s difficult to tell from this article whether they will massacre mankind or merely enslave them. One thing is for certain: there is no stopping them. The drones will soon be here.
And I for one welcome our new drone overlords. Iâ(TM)d like to remind them that as a trusted Slashdot personality, I can be helpful in rounding up others to toil in their underground battery caves.
But it sorta sounds like the guy looked at how video compression works (b-frames and whatnot) and realized he had a loosely analogous situation.
#DeleteChrome
Considering the payload capacity what happens when it runs out of ammo? -assuming it's used for such purposes.
In any case obstacle avoidance is imperative for anything moving by itself. I imagine there will be rigorous testing standards legislation for such capabilities in the future.
A 'singular oddity' is an event that cannot be explained and only happens when you are alone.
The video shows the detected obstacles, and the drone by far fails to detect all relevant obstacles. It looks like they rely on quite a bit of luck or they would crash into other obstacles while avoiding the obstacles they detect. Also, the world doesn't change much while you travel 10 meters, if you travel quickly, but your view of the world changes a lot when you turn. Each time they avoid an obstacle, the drone has not seen the nearest 10 meters in the new direction, and if an obstacle happened to be in the new path, it could very well already be too close to detect and avoid. The drone behaves very much like a panicky driver who veers into oncoming traffic to avoid hitting a squirrel.
Birds use a simpler approach: no 3d modeling; they just respond to relative speed of edges on their retina.
Already do that don't they? At least I do.. I tend to not worry too much about the first 10 meters in front of me because if something happens in that space I doubt I'd be able to avoid it!
Obviously the range at which I don't pay much attention changes with speed and location but the basic principle is sound.
The trick is to run out of obstacles before you run out of ammo.
What happens if something gets in its way after it has taken its picture. Or is that just and 'edge case'
Why limit the device? The two aren't mutually exclusive, so why not avoid some obstacles while blasting away others?
Guns don't kill people; Physics kills people! - John Lithgow as Dick Solomon on Third Rock From The Sun
I wonder how well it fares with moving obstacles... like others drones with the same 10 meters of horizon.
Here's the problem with this idea: it assumes that the ten meters in front of the drone won't change. That's not how the real world works. For instance, someone could stand up five meters in front of a low-flying drone using this methodology and get a drone in the face.
It is very reasonable to assign an overriding quality metric to a collision avoidance methodology by how fast it can react. That is one of the factors that should determine the maximum usable velocity of the drone. This method is very poor quality in just that way; so it's hardly a justification for a drone to travel at higher speeds.
I've fallen off your lawn, and I can't get up.
But if they approach an airplane at an angle so that the plane is always at the same bearing, it gets filtered out, the bird thinks the plane is at inifinity. Only when it is too near, the increased in perceived size of the object will create parallax. But by that time it is too late because the planes move too fast.
sed -e 's/Chuck Norris/Rajnikant/g' joke > fact
So it's only noticing if something is in the way when it gets within 10 meters?
30 mph = ~48 kph = ~13 mps. So it has less than a second to respond to whatever it does see. What's the turning radius for this thing when traveling at 30 mph? Or the stopping distance?
It's amusing that the thought line is "it's traveling so slow that i can do the processing to allow it to travel faster than many people would be comfortable with".
This Space Intentionally Left Blank
Or you could equip it with lasers!
so they created AI??? BIG NEWS