Design of an AI-based workflow-guiding system for stratified sampling

Abstract

The characterization of the resistance of transmission towers is a difficult and costly procedure which can be mitigated using statistical techniques. A stratified sampling process based on the characteristic of the terrain was shown in previous works to reduce the error in the statistical inference; however, such characteristics are usually unknown before a measure is made. In this work, we present a system which integrates artificial intelligence techniques, such as k-nearest neighbors, decision trees, or random forests, to automatically optimize the workflow of expert workers using various sources of data.