Talk:Termite-inspired robots

Blueprints or not?

The wikipedia article states that "No robot had a blueprint for constructing a particular structure".

The first published article (http://www.eecs.harvard.edu/ssr/papers/rss11-petersen.pdf) states that : "A user provides as input a high-level representation, specifying only where blocks should be located in the target structure. A simple offline compiler uses depth- first search to convert this to a “structpath” representation, a directed path in the plane annotated with the height of the stack of blocks at each site in the path"

And then, the pseudo code of the robots :

  climb onto structure
  while on structure do
     follow structpath
        if holding block then
           if structpath specifies block at current site

So the robot has the "structpath", which is a transformed version of the final structure.

How is this not a global vision of the end result? — Preceding unsigned comment added by 193.55.175.20 (talk) 16:30, 17 February 2014 (UTC)

• 193, I think you are correct. A structpath that is a directed path that is constructed from a high-level representation, which appears to be essentially a blueprint. Your point also raises a concern about how termite-inspired the robots are. Termites don't have anything like structpaths for building mounds. I suspect that that may be some criticism in the future, but now our own interpretations cannot be included. I am going to remove claims of no blueprints or plans since that is not consistent with what they did. I am One of Many (talk) 22:21, 17 February 2014 (UTC)