Full e-thesis link is provided below. If the link is broken, the thesis can also be found in the University of Nottingham e-Theses Repository.
Abstract:
With increasing emphasis on Industry 4.0 and 5.0 paradigms, there is interest in using data-driven technologies for improvements in the manufacturing sector. Personalisation is one such technology. Using personal data, this technology can change the appearance of a user interface or the action required by a user. This can make computer systems easier to use or better allocation of resources. For manufacturing industries, personalisation could be used to divert complex tasks to advanced users while allowing beginner users to learn on simpler operations. However, integrating data-driven technologies may produce unforeseen negative eGects. There is limited current knowledge of how personalisation works in the current manufacturing industry and how this may impact users. This thesis studies the potential eGect of personalisation when implemented in manufacturing scenarios.
The overarching goal of the thesis is to provide a set of design insights for personalisation’s usage in the manufacturing industry. An early collaborative work into Digital Manufacturing Technologies revealed the concerns around personal data capture and usage. This became a key motivator of the thesis. With personal data being a major component of personalisation, how people feel about data and therefore, the technology could be a potential barrier to its use in manufacturing settings. The current project’s aims focused on how people would design their own personalisation systems, understanding stakeholders’ acceptance of personalisation systems and how personalisation can aGect task completion.
To study a diverse range of topics regarding personalisation a mixed methods approach was utilised. Towards the start of the project, an examination of personalisation systems revealed a range of diGerent types. Thus, a taxonomy was defined, which became the underpinning for how the thesis treats personalisation. The proposed taxonomy categorises personalisation systems into five types whose name implies their function: Suggesters, Rearrangers, Swappers, Generators and Controllers. These categories became a type of card featured within constructed Personalisation Design Cards (PDCs). The PDCs were utilised to understand how users would design their own personalisation systems in a participatory design study. Personalisation’s acceptance from stakeholders was examined using a survey, with experimental conditions to represent data types used within the technology and diGerent personalisation types. The last empirical chapter featured multiple types of automated assistance within a quality control inspection task. The impact that the automation had on the participants and their feelings towards the assistance were captured. The types of personalisation system were subject to an additional systematic review at a later stage to provide further confirmation of their eGectiveness in categorising personalisation systems.
The empirical studies each contributed to the understanding of personalisation. The participants used PDCs were used to craft personalisation systems, this experience led to them describing similar worries about data from the motivating work. It appeared that the participants had a preference for systems that are “dynamic” and utilised less intrusive forms of data. The survey indicated that stakeholders preferred Swapper systems when the utilised personal data was heart rate but did not have a preference when performance data was used. Personalisation implementation within a Quality Control Inspection task demonstrated that users viewed the Controllers system (which “completed” the next action) as judgemental; in contrast, the Suggesters system (which suggested the next action) was recognised as having good usability. These insights were collated to form a set of design insights for industry professionals.
The thesis stands as an early look into personalisation’s usage within the manufacturing sector. It provides a resource for systems integrators which describes the theoretical underpinning of how this technology can be implemented. This enables potential integrators to select the type of personalisation that may be useful to them while simultaneously providing a baseline level of user requirements regarding how users may react to this technology. Future work can branch oG the included work by specialising in one aspect of personalisation, such as algorithms or design.
Faculty of Engineering
Supervisors: Robert Houghton, Ayse Kucukyilmaz, Sarah Sharples
Keywords: Data-driven technologies, Data-driven technologies, Manufacturing sector, Personal data, System design
Systems integrators
