Automatic detection and response to environmental change

Abstract

Robots typically have many sensors which are underutilized. This is usually because no simple mathematical models of the sensors have been developed or the sensors are too noisy to use techniques which require simple noise models. We propose to use these underutilized sensors to determine the state of the environment in which the robot is operating. Being able to identify the state of the environment allows the robot to adapt to current operating conditions and the actions of other agents. Adapting to current operating conditions makes the robot robust to changes in the environment by constantly adapting to the current conditions. This is useful for adapting to different lighting conditions or different flooring conditions amongst many other possible desirable adaptations. The strategy we propose for utilizing these sensors is to group sensor readings into statistical probability distributions and then compare the probability distributions to detect repeated states of the environment.

A Context-Based State Estimation Technique for Hybrid Systems

“…Context identification through classification is used in the computer vision community to enhance object recognition, and in the robotics community to adapt sensor parameters to the environment, assess the accessibility of a terrain and even build motion models [12], [13], [14], [15]. The essentially geometric classification effective in machine vision problems assumes that changes in signals are slow [12].…”

“…The essentially geometric classification effective in machine vision problems assumes that changes in signals are slow [12]. Classification of dynamic data is addressed by the robotics community, but under the assumption that the world can be described with a set of discrete, static states [13]. Techniques that do account for continuously changing states assume that the dynamics are time-invariant [15].…”

Bootstrap Learning and Visual Processing Management on Mobile Robots

“…There has been little work on color constancy in the presence of shadows and artifacts due to rapid camera motion. Further, with few exceptions (e.g., [41,42]), most methods do not function in real time.…”

An Illumination Identification System for the AIBO Robot

“…Lenser and Veloso [5] use a mathematical algorithm to detect changes in illumination over time. They implement a system that uses this detection algorithm to switch between several sets of pre-calibrated vision thresholds.…”