Functional Modeling for Monitoring of Robotic System

Haiyan Wu*, Rikke R. Bateman, Xinxin Zhang, Morten Lind

*Corresponding author for this work

Research output: Contribution to journalJournal articleResearchpeer-review

1 Downloads (Pure)

Abstract

With the expansion of robotic applications in the industrial domain, it is important that the robots can execute their tasks in a safe and reliable way. A monitoring system can be implemented to ensure the detection of abnormal situations of the robots and report the abnormality to their human supervisors or cooperators. In this work, we focus on developing a modeling framework for monitoring robotic system based on means-end analysis and the concept of action phases from action theory. A circular cascaded action phase structure is proposed for building the model of cyclical robotic events. This functional model provide a formal way of decompose robotic tasks and analyze each level of conditions for an action to be executed successfully. It can be used for monitoring robotic systems by checking the preconditions in the action phases and identifying the failure modes. The proposed method is demonstrated by using a simulated robotic manipulation system. The simulation results demonstrate the feasibility of the developed functional model in finding errors during the execution monitoring.
Original languageEnglish
JournalApplied Artificial Intelligence
Volume32
Issue number3
Pages (from-to)1-24
ISSN0883-9514
DOIs
Publication statusPublished - 2018

Cite this

@article{791b99398628412798652b1e01bba5ee,
title = "Functional Modeling for Monitoring of Robotic System",
abstract = "With the expansion of robotic applications in the industrial domain, it is important that the robots can execute their tasks in a safe and reliable way. A monitoring system can be implemented to ensure the detection of abnormal situations of the robots and report the abnormality to their human supervisors or cooperators. In this work, we focus on developing a modeling framework for monitoring robotic system based on means-end analysis and the concept of action phases from action theory. A circular cascaded action phase structure is proposed for building the model of cyclical robotic events. This functional model provide a formal way of decompose robotic tasks and analyze each level of conditions for an action to be executed successfully. It can be used for monitoring robotic systems by checking the preconditions in the action phases and identifying the failure modes. The proposed method is demonstrated by using a simulated robotic manipulation system. The simulation results demonstrate the feasibility of the developed functional model in finding errors during the execution monitoring.",
author = "Haiyan Wu and Bateman, {Rikke R.} and Xinxin Zhang and Morten Lind",
year = "2018",
doi = "10.1080/08839514.2018.1447431",
language = "English",
volume = "32",
pages = "1--24",
journal = "Applied Artificial Intelligence",
issn = "0883-9514",
publisher = "Taylor & Francis Inc.",
number = "3",

}

Functional Modeling for Monitoring of Robotic System. / Wu, Haiyan; Bateman, Rikke R.; Zhang, Xinxin; Lind, Morten.

In: Applied Artificial Intelligence, Vol. 32, No. 3, 2018, p. 1-24.

Research output: Contribution to journalJournal articleResearchpeer-review

TY - JOUR

T1 - Functional Modeling for Monitoring of Robotic System

AU - Wu, Haiyan

AU - Bateman, Rikke R.

AU - Zhang, Xinxin

AU - Lind, Morten

PY - 2018

Y1 - 2018

N2 - With the expansion of robotic applications in the industrial domain, it is important that the robots can execute their tasks in a safe and reliable way. A monitoring system can be implemented to ensure the detection of abnormal situations of the robots and report the abnormality to their human supervisors or cooperators. In this work, we focus on developing a modeling framework for monitoring robotic system based on means-end analysis and the concept of action phases from action theory. A circular cascaded action phase structure is proposed for building the model of cyclical robotic events. This functional model provide a formal way of decompose robotic tasks and analyze each level of conditions for an action to be executed successfully. It can be used for monitoring robotic systems by checking the preconditions in the action phases and identifying the failure modes. The proposed method is demonstrated by using a simulated robotic manipulation system. The simulation results demonstrate the feasibility of the developed functional model in finding errors during the execution monitoring.

AB - With the expansion of robotic applications in the industrial domain, it is important that the robots can execute their tasks in a safe and reliable way. A monitoring system can be implemented to ensure the detection of abnormal situations of the robots and report the abnormality to their human supervisors or cooperators. In this work, we focus on developing a modeling framework for monitoring robotic system based on means-end analysis and the concept of action phases from action theory. A circular cascaded action phase structure is proposed for building the model of cyclical robotic events. This functional model provide a formal way of decompose robotic tasks and analyze each level of conditions for an action to be executed successfully. It can be used for monitoring robotic systems by checking the preconditions in the action phases and identifying the failure modes. The proposed method is demonstrated by using a simulated robotic manipulation system. The simulation results demonstrate the feasibility of the developed functional model in finding errors during the execution monitoring.

U2 - 10.1080/08839514.2018.1447431

DO - 10.1080/08839514.2018.1447431

M3 - Journal article

VL - 32

SP - 1

EP - 24

JO - Applied Artificial Intelligence

JF - Applied Artificial Intelligence

SN - 0883-9514

IS - 3

ER -