Detecting Performance Anomalies in the Multi-component Software a Collaborative Robot

• Héctor Quintián ^[1] ; Esteban Jove ^[1] ; José Luis Calvo-Rolle ^[1] ; Nuño Basurto ^[2] ; Carlos Cambra ^[2] ; Álvaro Herrero ^[2] ; Emilio Corchado ^[3]
  1. [1] Universidade da Coruña

     Universidade da Coruña

     A Coruña, España

     
  2. [2] Universidad de Burgos

     Universidad de Burgos

     Burgos, España

     
  3. [3] Universidad de Salamanca

     Universidad de Salamanca

     Salamanca, España

     

  Mostrar afiliaciones +
• Localización: Intelligent Data Engineering and Automated Learning – IDEAL 2020. 21st International Conference: Guimarães, Portugal; November 4–6, 2020. Proceedings / Cesar Analide (ed. lit.), Paulo Novais (ed. lit.) , David Camacho Fernández (ed. lit.) , Hujun Yin (ed. lit.), Vol. 2, 2020 (Part II), ISBN 978-3-030-62365-4, págs. 533-540
• Idioma: inglés
• Enlaces
  • Texto completo
• Resumen

  • The detection of anomalies (affecting hardware or software) is an open challenge for cyber-physical systems in general and robots in particular. Physical anomalies related to the hardware components of such systems have been widely researched. However, scant attention has been devoted so far to study the anomalies affecting the software components. In order to bridge this gap, the present paper proposes the application of different classifiers to a robot performance dataset for the first time. The applied supervised models are targeted at detecting synthetically-induced software anomalies, having a detrimental impact on the performance of a collaborative robot. Obtained results demonstrate that the applied Machine Learning models can successfully address the target problem, with acceptable detection rates.