Back to search

Democritus University of Thrace (DUTH)

System safety approaches for human-robot collaborative tasks

Abstract

dc:description

Τhis dissertation aims to provide a real-time decision-making mechanism tool, which capitalizes on the determined unsafe control actions extracted from the STPA, in order to prompt the system into the selection of an action, based on the current and past observations and actions, that will transit it to a next state that belongs to a higher safety level. Therefore, the proposed HRC safety method is targeted as a step forward towards a generic safe HRC framework for various robotic platforms, which perform collaborative tasks with human operators. We employ an abstraction for clustering the plethora of the normal and unsafe states of the system, and we formulate the problem as a partial observable Markov decision process to determine a policy graph that operates in real-time and deduces the safe and unsafe workflows of the system. Thus, this dissertation provides a theoretical framework for reliable and safe HRC in unconstrained environments and shows great potential in practical applications. Moreover, the proposed safety method offers a new perspective to view and implement safe HRC by combining various methodologies in different research fields, including observation theory, graph policy, systems, and control theory.

Degree

thesis:*
Grantor dc:publisher
Democritus University of Thrace (DUTH)
Year dc:date
2023

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Zacharaki, Angeliki
  • Ζαχαράκη, Αγγελική

Subjects

dc:subject × 20

Rights

Language dc:language
eng

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/53599

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Zacharaki, Angeliki; Ζαχαράκη, Αγγελική. System safety approaches for human-robot collaborative tasks. Democritus University of Thrace (DUTH), 2023. http://hdl.handle.net/10442/hedi/53599