Navigating in 3-d using Human-Level Artificial Intelligence
(pt 2)
Note: To make this website free to the public
please click on an ad to support my sponsors or you
can make a tax-deductable donation using Paypal
(click on the donation icon on the left).
This video shows a robot
navigating in a 3-dimensional environment. There are
no sound in this video because I wanted the viewers
to focus on the robot's thoughts while navigating in
a 3-d plane.
When navigation
questions are asked, such as: "where is the nearest
library" or "what did you see inside the super
market", the robot has to have a 3-d map of
everything it has explored in the past. The robot
will recall vital experiences and extract facts from
these experiences.
I have broken up this
lesson into two parts. 1. navigation in a maze. 2.
answering questions based on your navigation. In the
first part, I try to show how the robot stores
experiences in memory regarding navigation. The robot
stores visual experiences, frame-by-frame, in memory
and these frames are stored in memory in a 3-d manner
(each frame has a direction and a link to the next
frame).
In this video, the
robot has to navigate through an unknown mansion
(maze). The instructions state: the player has to
find the main door, search for money and defeat
zombies. At the end, the player does indeed find the
main door, but has trouble finding a solution to open
the main door.
The first video is
about how does the robot form a 3-d map in memory on
the current environment and how does the robot use
this 3-d map to navigate in the maze. here is a link
to the first video:
http://www.youtube.com/watch?v=OWKJamRVySw
Keep in mind that all
knowledge to navigate in a 2-d or 3-d world learned
from teachers in school. Teachers teach the robot how
to walk from place to place or to do maze worksheets
(2-d navigation). No predefined navigation system is
used. No 3-d mapping software is used.