The driving hypothesis of this thesis is that a quantitative approach linking business objectives of an organization with technological limitations of the physical product would enable industry to create more innovative products. The main goal of this research is to validate the applicability and reliability of the innovation mining framework developed by Peyyeti (2016) to identify innovation opportunities and components worth innovating in a product. In this work, the innovation mining framework is applied with minor modifications to a mechanical pencil, innovation scenarios were then compared to existing innovations in mechanical pencils. Based on the success of the feasibility trial, the innovation mining framework was applied to a Dirt-Devil vacuum and compared to innovations implemented in the Dyson-V6 vacuum to improve a set of chosen value-metrics. Based on this study, the following insights were developed: (1) The model sufficiently identified several innovation opportunities to improve each value-metric (2) Varying weighting schemes do not have significant effects on filtered data (3) The top-half of the dendrogram contains the most relevant clusters that present viable innovation opportunities (4) The relevant clusters must be viewed from a systems thinking perspective as a single chain that must be innovated for the most benefit (5) Implementing this model provokes systems thinking approach in the user. This gives a substantial advantage over intuitive and qualitative approaches by providing insights on hidden relationships and identifying innovation opportunities in a system that may otherwise be ignored or unexplored. Opportunities for future-work include developing a transfer-function system representing true relationships, performing SVD at every level of the coupling matrices to gain insights into the nature of transformation and cluster formation, comparing clusters obtained to failure-modes associated with the corresponding value-metric for systematic prioritization and comparing dendrogram clusters with function-structure map to get detailed insights on clusters and their interactions.
Industrial and Systems Engineering (MS)
Department, Program, or Center
Industrial and Systems Engineering (KGCOE)
Subramani, Karthikeyan, "Validation of an Innovation Mining Framework" (2018). Thesis. Rochester Institute of Technology. Accessed from
RIT – Main Campus