Self-Certification and Safety Compliance for Robotics Platforms

Osama Zaki, David Flynn, Jamie Blanche, Joshua Roe, Chi Wai (Leo) Kong, Daniel Mitchell, Theodore Lim, Sam Harper, Valentin Robu

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

In this paper, the results and methodology of a framework to enable run-time safety compliance and self-certification of robotics is presented. This transferable framework is verified within a practical demonstration scenario, based on asset inspection within a confined space, and representing a Beyond Visual Line of Sight (BVLOS) use case. The methodology of the framework is based on computationally efficient analysis to support run-time, front-end, data analysis and adaptive decision-making. Utilizing the Husky A200 platform, manufactured by Clearpath, front-end datasets on the mission status and diagnostics of critical sub-systems within the Husky platform are used to update run-time system ontologies. The holistic and hierarchical- relational model of the robot integrates the automata of the sensed and some non-sensed components, using prior knowledge, such as risk assessments and offline reliability data, to support run-time analysis, such as fault prognosis, detection, isolation and diagnosis. These computationally efficient data and system analyses then enable faults to be translated into failure modes that can affect decision making during the mission. With respect to challenges of a dynamic environment, namely ambient conditions or the presence of unexpected people, Frequency Modulated Continuous Wave (FMCW) sensing is integrated onto the husky platform. The FMCW supports localization in opaque environments and detection of people within and out-with of the confined space, as well as enabling integrity analysis of the infrastructure. The framework presents its results within a symbiotic digital twin of the infrastructure and robotic platform. With fully synchronized communication and data streams, the interactive digital twin provides operational decision support and trust for human in the loop operators of varying skill levels. The presentation of actionable information to the end user is used to support improvements in productivity associated with asset integrity as well as supporting user trust in safety during a BVLOS mission.
Original languageEnglish
Title of host publicationOffshore Technology Conference 2020
Number of pages18
ISBN (Electronic)978-1-61399-707-9
DOIs
Publication statusPublished - 4 May 2020
EventOffshore Technology Conference 2020 - Houston, United States
Duration: 4 May 20207 May 2020
http://2020.otcnet.org/

Conference

ConferenceOffshore Technology Conference 2020
Abbreviated titleOTC 2020
CountryUnited States
CityHouston
Period4/05/207/05/20
Internet address

ASJC Scopus subject areas

  • Safety, Risk, Reliability and Quality
  • Ocean Engineering
  • Energy Engineering and Power Technology
  • Mechanical Engineering

Fingerprint Dive into the research topics of 'Self-Certification and Safety Compliance for Robotics Platforms'. Together they form a unique fingerprint.

Cite this