Analysis of technological knowledge stock and prediction of its future development potential: The case of lithium-ion batteries

Amidst growing environmental challenges and increasing penetration of sustainable energy production, the development of efficient energy storage system (EES) is indispensable to fully benefit from the use of intermittent renewable energy sources. Although lithium-ion batteries (LIBs) have been the dominant energy storage technology for applications in consumer electronics, electrified transportation and stationary electricity storage, further technological progress is required for achieving the climate protection objectives. In this context, continuous exploration of new technological opportunities is favorable to advance the global energy transition towards sustainability to success. To facilitate such progress and to avoid potential technological lock-in situation, understanding the underlying latent knowledge construct revealed in patents may prove beneficial. However, previous studies have either focused on analyzing the technological improvements along a particular development path or emphasizing the necessity for improvements in environmental-economic dimensions. Accordingly, a novel patent-based analysis framework is proposed to obtain an inclusive view on the continuing path of technology development concerning the research of LIBs as well as to predict its future development potential based on two machine learning algorithms: Principal component analysis (PCA) and random forest classifier (RFC). In this study, PCA is firstly applied to group various knowledge areas into key technological knowledge stocks. Building upon these results, RFC is adopted to forecast the future developmental potential of technological knowledge areas. The findings show that the research landscape of LIBs is a dynamic environment where new knowledge stocks emerged overtime and the identified knowledge stocks within a defined time interval play an orchestrating role in the advancement of lithium-ion battery technology. In particular, knowledge stocks related to battery management system and hybrid capacitors have gained relevance. The proposed analysis framework can be used to track the past chronology of knowledge accumulation trajectory and can proactively enhance the knowledge management capabilities of experts, who wish to create a path towards a low-carbon economy. (C) 2019 Elsevier Ltd. All rights reserved.