Roadmap

From TinyCog
Jump to: navigation, search

The current release is TinyCog 0.0.1. This release include the following running examples for testing Scenes and basic Description Logic:

  • test010-dl-family-prim.pl - Description Logic "Primitive subsumption" test
  • test040-hanoi-noplan.pl - Hanoi towers with a hard-coded planner example

However, the other tests still include errors.

TinyCog 0.0.2

This release should fix the current issues and should get the following examples running:

TinyCog 0.0.3

The next step is to get the Hunter Domain running, allowing a human operator to inspect the agent's state and to give basic orders. This domain is based on basic urges of the agents (i.e. being hungry) and some hard-coded task decompositions for the planner in order to deal with the urges (grazing or hunting).

After that, we would introduce social collaboration for group hunting and an Episodic Memory in order to collect statistics about the hunting success.

TinyCog 0.0.4

We don't know what comes next after that, this really depends on _your_ comments here on the forum. But we know where we want to end up in the far future: with a TinyCog as a module for [ROS (Robot Operating System)] controlling a robot that is capable of understanding and interacting with humans and other robots.