Mining Patent Precedents for Data-Driven Design: The Case of Spherical Rolling Robots

Abstract

Data-driven engineering designers often search for design precedents in patent databases to learn about relevant prior arts, seek design inspiration, or assess the novelty of their own new inventions. However, patent retrieval relevant to the design of a specific product or technology is often unstructured and unguided, and the resultant patents do not sufficiently or accurately capture the prior design knowledge base. This paper proposes an iterative and heuristic methodology to comprehensively search for patents as precedents of the design of a specific technology or product for data-driven design. The patent retrieval methodology integrates the mining of patent texts, citation relationships, and inventor information to identify relevant patents; particularly, the search keyword set, citation network, and inventor set are expanded through the designer's heuristic learning from the patents identified in prior iterations. The method relaxes the requirement for initial search keywords while improving patent retrieval completeness and accuracy. We apply the method to identify self-propelled spherical rolling robot (SPSRRs) patents. Furthermore, we present two approaches to further integrate, systemize, visualize, and make sense of the design information in the retrieved patent data for exploring new design opportunities. Our research contributes to patent data-driven design.

Design-by-Analogy: Exploring for Analogical Inspiration With Behavior, Material, and Component-Based Structural Representation of Patent Databases

“…To address this challenge, many computational tools and methodologies have been studied. Song and Luo integrated the mining of patent texts, citation relationships, and inventor information to retrieve patents for assisting data-driven design [26]. Murphy used a Vector Space Model algorithm to evaluate functional analogies within patents [3].…”

Engineering Knowledge Graph for Keyword Discovery in Patent Search

“…In general, COM is suitable for patent retrieval for general technological domains, such as nanotechnologies, fuel cells, and electric motors. In addition, Song and Luo (2017) proposed a method to integrate the keyword search in patent texts and the searches using citation relationship and co-inventor relationship between patents, and use an iterative process to read the previously retrieved patents to expand the keyword queries, in order to arrive at a more comprehensive and accurate search. Another strand of approaches exploits semantic analysis to search for patents in the vast databases.…”

“…An of example SRR is presented in Figure 4. An SRR patent set has been reported in a recent study (Song and Luo, 2017), which aimed to exhaustively retrieve patents related to SRRs from the USPTO database. Herein, we use the same case because of our rich experience in SRR patent search and for the convenient validation of the new EKG-based search keyword discovery and patent retrieval method.…”

“…It is a non-trivial task to retrieve the set of patents that are accurately relevant to a specific search interest, such as 3D printing and rolling robots, for the interests of engineers, designers, and innovators. The existing retrieval methods rely on keyword search in patent texts (Fujii, 2007;Koch et al, 2011;Nakamura et al, 2015) and mining additional structured information in patent documents, such as inventors, assignees, classification and citations (Wang, 2011;Benson and Magee, 2013;Song and Luo, 2017). Keyword search in patent texts is the key in general patent retrieval procedures, methods, and tools, but the keywords are often intuitively chosen by searchers subject to their knowledge and vocabulary on the technology of search interest (Fujii, 2007;Wang, 2011;Montecchi, Russo and Liu, 2013).…”

Predicting Part Mass, Required Support Material, and Build Time via Autoencoded Voxel Patterns

“…However, we are not aware of any attempt to utilize these algorithms to provide design-centric feedback to assist novice designers. More generally, research on data-driven methodologies for engineering design is on the rise [32] and has included text-based design descriptions [33,34], patent text [35][36][37], and even online reviews [38,39]. However, the rise of online communities for solid modeling and engineering design (such as GrabCAD and Thingiverse) has made entirely new sources of data available for data mining and machine learning approaches.…”