A Conceptual Model

Conceptually, a cognitive system should have

  1. Components to process the signals from the sensors, creating percepts.
  2. A knowledge or belief base, containing what the agent knows or believes.
  3. Methods for maintaining the belief base, given the stream of incoming percepts.
  4. A way to represent the agent’s knowledge in a compact form, which is amenable for use in reasoning and planning.
  5. A set of models for how the environment works (i.e., environment dynamics).
  6. Reasoning capabilities, i.e., inference for query answering.
  7. Planning capabilities, i.e., generate recommendations of sequences of actions.
  8. A way of executing of actions such that they do not conflict with each other and are executed in acceptable time limits.
  9. The ability to learn and adapt.
  10. The ability to deal with uncertainty (noise).
  11. The ability to deal with ignorance (lack of knowledge).

One architecture we can find inspiration from is the HPB agent architecture (Rens & Meyer, 2015; Rens & Moodley, 2017). Here is a flow-diagram representing the architecture:

 

Another architecture we can find inspiration from is the Stochastic Belief Management framework (Rens, Meyer & Moodley, 2017). Here is a high-level conceptual model:

A more practical architecture for a computational cognitive system might be composed of six main modular components.

  1. Signal processing and observation buffering.
  2. A probabilistic belief base and belief change operators.
  3. Belief state condensation.
  4. Environment dynamics representation.
  5. Intention and goal management.
  6. Real-time stochastic action planning under partial observability.

Each module would be implemented with Artificial Intelligence techniques and algorithms from the state of the art.