article · 2015
An Embedded Implementation of Bayesian Network Robot Programming Methods
Cited by 1Open access
Abstract
A wide variety of approaches exist for dealing with uncertainty in robotic reasoning, but relatively few consider the propagation of statistical information throughout an entire robotic system. The concept of Bayesian Robot Programming involves making decisions based on inference into probability distributions, but can be complex and difficult to implement due to the number of priors and random variables involved. In this work, we apply Bayesian network structures to a modified Bayesian programming paradigm to provide intuitive structure and simplify the programming process. The use of discrete random variables in the network can allow high inference speeds, and an efficient programming toolkit suitable for use on embedded platforms has been developed for use on mobile robots. A simple example of navigational reasoning for a small mobile robot is provided as an example of how such a network can be used for decisional programming.
Cite this paper
Post, M. (2015). An Embedded Implementation of Bayesian Network Robot Programming Methods. https://doi.org/10.19124/ima.2015.001.06
Post, Matt. “An Embedded Implementation of Bayesian Network Robot Programming Methods.” 2015, https://doi.org/10.19124/ima.2015.001.06.
Post, Matt. 2015. “An Embedded Implementation of Bayesian Network Robot Programming Methods.” https://doi.org/10.19124/ima.2015.001.06.
Post, M. (2015) ‘An Embedded Implementation of Bayesian Network Robot Programming Methods’. Available at: https://doi.org/10.19124/ima.2015.001.06.
@article{post2015embedded,
title = {{An Embedded Implementation of Bayesian Network Robot Programming Methods}},
author = {Post, Matt},
year = {2015},
doi = {10.19124/ima.2015.001.06}
}Made from the index record. Check names, capitalisation and page numbers against the paper. Need another style? AI for Research has over 10,000.