Companies make substantial R&D investments in early design stages to develop radically innovative products. However, despite abundant research work in the field of human-centered design, the front end of innovation is the least well-structured part of the innovation process. Radical Innovation Design (RID) methodology has put forward a structured process aiming at exploring, organizing and categorizing required knowledge to design a useful problem in the form of value buckets (i.e. overlooked problems of users) to be evaluated. In this paper, we aim at reinforcing the knowledge acquirement process in parallel with the problem design process to fine-tune the firm's R&D strategies, and to increase the likelihood of successfully reaching the mainstream markets. The contributions of this paper are twofold: i) to provide a set of knowledge acquisition rules in front end of innovation; and ii) to specify a set of problem design guidelines mainly by introducing a tool called Dependency Structure Modeling (DSM)-Value bucket algorithm, which enables a systematic identification of value creation opportunities. The example of a smart lighting project is analyzed in this paper as a case study.