Robust multi-objective optimization for solving the RFID network planning problem

Abstract

Radio-frequency identification (RFID) is a new technology used for identifying and tracking objects or people by radio-frequency waves to facilitate automated traceability and data collection. The RFID system consists of an electronic tag attached to an object, readers, and a middleware. In the latest real applications based on the RFID technology, the deployment of readers has become a central issue for RFID network planning by means of optimizing several objectives such as the coverage of tags, the number of readers, and the readers/tags interferences. In practice, the system is affected by uncertainty and uncontrollable environmental parameters. Therefore, the optimal solutions to the RFID network planning problem can be significantly reduced with uncontrollable variations in some parameters, such as the reader's transmitted power. In this work, we propose a robust multi-objective optimization approach to solve the deployment of RFID readers. In this way, we achieve robust optimal solutions that are insensitive to uncertainties in the optimization parameters.

A generic model of the information and decisional chain using Machine Learning based assistance in a manufacturing context

“…Machine learning techniques have obtained remarkable achievements in various tasks, such as image recognition, object detection, and language modeling. However, building a high-quality ML system for a specific task highly relies on human expertise, hindering its wide application [45,46]. Meanwhile, automated machine learning (AutoML) is a promising solution for building a ML system without human intervention.…”