Thesis-2012-Fu.pdf (2.64 MB)
A model for assessment of human assistive robot capability
thesis
posted on 2012-06-28, 11:13 authored by Huazhong FuThe purpose of this research is to develop a generalised model for levels of autonomy and sophistication for autonomous systems. It begins with an introduction to the research, its aims and objectives before a detailed review of related literature is presented as it pertains to the subject matter and the methodology used in the research. The research tasks are carried out using appropriate methods including literature reviews, case studies and semi-structured interviews.
Through identifying the gaps in the current work on human assistive robots, a generalised model for assessing levels of autonomy and sophistication for human assistive robots (ALFHAR) is created through logical modelling, semi-structured interview methods and case studies. A web-based tool for the ALFHAR model is also created to support the model application. The ALFHAR model evaluates levels of autonomy and sophistication with regard to the decision making, interaction, and mechanical ability aspects of human assistive robots. The verification of the model is achieved by analysing evaluation results from the web-based tool and ALFHAR model. The model is validated using a set of tests with stakeholders participation through the conduction of a case study using the web-based tool.
The main finding from this research is that the ALFHAR model can be considered as a model to be used in the evaluation of levels of autonomy and sophistication for human assistive robots. It can also prove helpful as part of through life management support for autonomous systems. The thesis concludes with a critical review of the research and some recommendations for further research.
History
School
- Mechanical, Electrical and Manufacturing Engineering
Publisher
© Huazhong FuPublication date
2012Notes
A Doctoral Thesis. Submitted in partial fulfillment of the requirements for the award of Doctor of Philosophy of Loughborough University.EThOS Persistent ID
uk.bl.ethos.587959Language
- en