The sensori-motor learning experiments will comprise the comparison of successful 2-dimensional khepera like agents doing phototaxis on four different conditions:
A: fixed sensors ontogenitically and evolutionary
B: random sensors evolutionary – fixed ontogenetically.
C: fixed sensors evolutionary – random onotogenetic growth
D: random evolutionary placement & random ontogenetic growth
In each of them, the comparisons are in terms of how do the evolved agents cope with radical perturbations. In neither of them the agents will have any homeostatic learning mechanism, weight changing, etc. The motivation is to find out which of these conditions favours the maintanence of plasticity througout the lifetime of the agent so as to produce adaptations to sensori-motor perturbations, including radical ones (e.g. human adaptation to up/down inversion prism goggles).