Author Topic: Suggestion for Research topic  (Read 3601 times)

0 Members and 1 Guest are viewing this topic.

Offline ameerhamzakhanTopic starter

  • Beginner
  • *
  • Posts: 1
  • Helpful? 0
Suggestion for Research topic
« on: September 20, 2015, 02:01:01 AM »
Hi Guys,

I am looking for a topic to pursue doctorate level research. I have just completed my undergraduate studies and got a chance to be admitted in PhD, based on an excellent proposal for research? I have no prior in-depth background of research so it is difficult for me to come up with something entirely new. My main interest lies in swarm robotics and AI. Can you give any suggestion?

Offline mklrobo

  • Supreme Robot
  • *****
  • Posts: 558
  • Helpful? 15
  • From Dream to Design at the speed of Imagination!
Re: Suggestion for Research topic
« Reply #1 on: September 20, 2015, 01:15:34 PM »
 ;D Hello! I have 2 ideas, one abstract, another practical Ai based.
1> AI structure;
The basic AI is built up of 3 working codes; the commander, the recorder, and the Ex-O.
These 3 together make up a whole "droid" AI. 256 droids would inhabit a sector on a hard drive, working together. 1/2 of the droids were the work force, their "minds" in habiting droid bodies, while their "soul" was on the hard drive. This way, if a droid body was destroyed, another body was activated, and
took over the work that the other body was doing. This way, no work
data was lost, nor the "droid" destroyed. Another 1/4 of the "droids" were the Dr.'s, monitoring and surveying the working robots, repairing any problems that arose.
The last 1/4 of the "droids" are running the mission in a simulator. They
make sure the rest of the droids are following their mission. They also report
to the "collective" other disks, to join in on what the total collective is doing.
If any parts of the mission of the working "droids" are jeopardized, the
problem is reported to the collective, and a problem solution is generated.
Using this method, an infinity amount of "droids" would be generated to
solve the problem, assuming the hardware is available. I learned something
incredible while working on this; a single AI can be in several bodies at one time, relative to us.
(Like a scheduler in the Linux Kernel) Simply put, it is like multiplying yourself, and your work potential.
2> Application level.
In the movie, Silent Running, there were three robots, Huey, Duey, and Louey. One possible purpose for the droid army, is to maintain a garden. I was thinking about the Army maintaining a garden in my back yard, and maintaining an aquaponics system that I would have. The Axon could be used to navigate/control a farms combines, or tractors.
Good Luck!!!   ;) :) :D ;D

Offline mklrobo

  • Supreme Robot
  • *****
  • Posts: 558
  • Helpful? 15
  • From Dream to Design at the speed of Imagination!
Re: Suggestion for Research topic
« Reply #2 on: November 22, 2015, 09:26:29 AM »
 ;) Hello!
 Your AI swarm direction you are going in. RoboPi came up with a great idea! Use a Field Programmable Gated Array for Neural networks, and rename them FPNN;Field Programmable Neural Network. Basically, a whole "brain" on a chip!! Is that Wild, or what?!?  :o
The problem is, what "class" of robot body will the FPNN control?   ???

 


data_list