Re: I laugh when I hear people say "Amazon Air will never wo
MadMitch88 said:
Khudson7 said:
Drones will just need more detect and avoid sensors and autonomous flying like what is being developed at MIT in this video:
http://youtu.be/kYs215TgI7c
That video link will anger a lot of people in here, especially the ones who insist all commercial drones will need to be $50,000+ machines with massively complex collision-avoidance systems in order to be viable.
And yet MIT demonstrated a little foam flyer that can perform very accurate autonomous missions with cheap parts you'd find in a $200 Netbook.
Fascinating video, I am amazed by its flight around the obstacles....Thank you for sharing. However they state that this is a 2D environment only. A 3D space is obviously more difficult, which is 360 degrees in all directions, both vertical and horizontal and in-between. With current technologies, the input alone to travel the 3D space would have to be an exponential number of inputs for the now seemingly simple 2D space.
Further, they state the following.
Because the problem of autonomous plane navigation in confined spaces is so difficult, and because it’s such a new area of research, the MIT team is initially giving its plane a leg up by providing it with an accurate digital map of its environment. That’s something that the helicopters in the AUVSI challenges don’t have: They have to build a map as they go.
The MIT researchers’ next step will be to develop algorithms that can build a map of the plane’s environment on the fly. Roy says that the addition of visual information to the rangefinder’s measurements and the inertial data could make the problem more tractable. “There are definitely significant challenges to be solved,” Bry says. “But I think that it’s certainly possible.”
Further the rules state for the competition:
Computational power need not be carried by the air vehicle.
See Competition and Rules:
http://www.aerialroboticscompetition.org/
I believe we have some amazing people working on the subject, all the research institutes. I’ve seen some amazing feats of engineering by these people. If it to be solved, these people in research are the teachers of tomorrows answers.
Reading the fine print is required to bring everything into perspective. Computational power is the limiting factor. And yes, AI is loosely used today, just like how the word drone is used. With AI, you'll be able to create an algorithm that can utilize more "senses" at varying degrees; just like we humans do. More hearing than seeing when it's dark outside, as an example. This will reduce the computational demand. But again relies on AI, which is computations or algorithms, to determine when and where these sense's should heighten to reduce the computation overhead. I call it sensory overload when I feel I can't think straight, computers have 1 billionth of our computation powers. Research is looking more at nature to provide these autonomous answers, they are simpler machines that navigate a changing environment. Even swarms, flocks, herds and schools of animals are defining how we move forward on the subject, shared computational power via a system of multiple sensors on multiple devices.
If we are talking in the context of a centralized drop-off point, so many feet in the air, then we can have commercial deliveries starting tomorrow. Which would be good in so many respects and a simpler answer to the autonomous only problems. But that's not the conversation is it?
I by no means am saying it can't be done. The difficulties are great. I enjoy the conversation, everyone has great ideas that should be considered. Keep discussing, so we all may find answers that progresses the subject to a solution....