One-of-a-Kind Product Development for Mass Personalisation in a Cloud-Based Environment

Abstract

Nowadays, with the rapid development of information and communication technologies (e.g. Web 2.0, cloud computing, wireless sensor network) and manufacturing technologies (e.g. additive manufacturing), users have become more actively involved in the product development stage to create personalised products with higher efficiency. This emerging manufacturing paradigm is known as mass personalisation (MP), of which eliciting and analysing individual customer’s explicit and latent needs, or customer requirement (CR) management, is deemed the core. Co-creation (e.g. user participation), product change (e.g. adaptability of design and product), and user experience (UX) (e.g. emotional factors and product utility) are regarded as key characteristics. Although one-of-a-kind production (OKP) companies, mostly small-and-medium-sized enterprises (SMEs), strive to deliver personalised product to achieve individual customer satisfaction, they still face many challenges and lack a systematic way to handle the high product variety, active user involvement, and better UX cost-effectively. Aiming to fill this gap, and also inspired by the prominent advantages of the cloud, this research focuses on formulating a systematic framework with novel methods to enable oneof- a-kind product development (OKPD) for MP in a cloud-based environment. For description consistency, the main content of this work can be further categorised into four aspects corresponding with the aforementioned core, and the three key characteristics of MP, in the two sequential stages of product development (i.e. product planning stage and product design stage) as follows: 1) A cloud-based co-creation platform for conducting the one-of-a-kind product planning (OKPP) process. Its business model, information communication process, and system architecture are proposed, which provide a promising means of SMEs to acquire abundant user information and to utilise various design resources, with little upfront capital investment. This platform is considered the basis of this thesis, and it also provides a potential co-creation environment to enable OKPP for MP. To validate it, the proposed cocreation platform is implemented in the Amazon EC2 cloud-based environment, illustrated by an example of a personalised bicycle product planning. 2) Novel customer-preference modelling and evaluation methods for CR management in the OKPP stage. CR management, often conducted in the early product development process, plays a significant role in achieving final product success. Based on the systematic procedures of the OKPP process, a weighted-preference graph method is proposed to handle vague and incomplete CRs in customer preference analysis. Moreover, a rough setbased fuzzy axiomatic design approach, that is, triangular rough number method, is proposed for the personalised product multi-attribute decision-making process. An illustrative example with a comparative study was undertaken thereafter to validate the feasibility and advantages of the proposed methods. Owing to their capability of dealing with limited CR information in the new product development process, therefore, they together provide a generic approach of managing CRs in OKPP for MP. 3) A bi-level two-stage design process, that is, modular design stage at macro level and scalable design at micro level, to enable product change of OKPD for MP. Based on the proposed cloud-based co-creation platform, both a personalised product configuration system, that is, adaptable technical configurator with open architecture products, and an embedded open toolkit, that is, a data-driven cyber-physical model, were established following the proposed two-stage design process. For the prior one, its system framework development and implementation process in the cloud-based environment are described in detail, with an illustrative example of a personalised bicycle design. The latter one extends the scope of the online product configuration system, which only deals with cyber models, that is, CAD models, by taking physical models (i.e. product with ICT-embedded components) into overall consideration. Accordingly, personalised smart, connected product, as a state-of-the-art case of personalised products was further discussed, with a case study of a personalised smart respiratory mask design. 4) Elicitation of UX in the personalised product co-development context. Apart from the above work, UX in the co-design content is also significant. A three-model-based, that is, physical model, cyber model and UX model, generic framework for conducting UX-based product co-development for MP is proposed, in which the three key characteristics of MP are considered collectively. To validate it, a 3-round human participation experiment was also carried out by developing a personalised smart wearable, that is, i-BRE respiratory mask. Both quantitative methods (e.g. sensor data recording and analysis, configurable components selection) and qualitative methods (e.g. a heat map of eye-tracking, & online questionnaire) were adopted to analyse individual UX in both a tangible (i.e. physical product prototype with smart functions) and visualised way (i.e. virtual model and configuration toolkits), coherently. Most of the research work in this thesis has been reported in several research publications, that is, six journal papers, one book chapter and four conference papers. It is hoped that the research outcome will offer some state-of-the-art knowledge and useful guidance to both product designers and OKP companies in their product development for MP.