Learning agent's spatial configuration from sensorimotor invariants

TL;DR

This paper explores how a robot can autonomously develop a spatial understanding from sensorimotor experiences without prior assumptions, demonstrating that the configuration space of its sensors can be learned through invariants in sensorimotor functions.

Contribution

It introduces a method for robots to learn spatial configurations from sensorimotor invariants, showing that environment-independent spatial notions emerge from motor-sensory mappings.

Findings

Robot learns the configuration space of its sensor system.

The learned manifold has the topology of a plane and a circle.

Spatial notions can be derived from sensorimotor invariants without prior assumptions.

Abstract

The design of robotic systems is largely dictated by our purely human intuition about how we perceive the world. This intuition has been proven incorrect with regard to a number of critical issues, such as visual change blindness. In order to develop truly autonomous robots, we must step away from this intuition and let robotic agents develop their own way of perceiving. The robot should start from scratch and gradually develop perceptual notions, under no prior assumptions, exclusively by looking into its sensorimotor experience and identifying repetitive patterns and invariants. One of the most fundamental perceptual notions, space, cannot be an exception to this requirement. In this paper we look into the prerequisites for the emergence of simplified spatial notions on the basis of a robot's sensorimotor flow. We show that the notion of space as environment-independent cannot be deduced solely from exteroceptive information, which is highly variable and is mainly determined by the contents of the environment. The environment-independent definition of space can be approached by looking into the functions that link the motor commands to changes in exteroceptive inputs. In a sufficiently rich environment, the kernels of these functions correspond uniquely to the spatial configuration of the agent's exteroceptors. We simulate a redundant robotic arm with a retina installed at its end-point and show how this agent can learn the configuration space of its retina. The resulting manifold has the topology of the Cartesian product of a plane and a circle, and corresponds to the planar position and orientation of the retina.