Advancing engineering R&D capabilities via text analytics

This R&D and engineering company was able to exploit big data analytics (BDA) to take a short-cut in an otherwise time-consuming and resource-intensive research and development process for high-tech products.  With advanced text analytics, internationally-filed patents can be quickly reviewed and the R&D team will always be informed of the latest technological advancements and market niches. Thus the company will always be up-to-date and fully aware of their competitors’ technological directions.

Challenges:
The R&D process for high-tech products is lengthy and resource-intensive. There are many tasks that require a lot of man-hours to complete. One of them is the studying of internationally filed patents, for two key objectives. First, the department will always be informed of the latest technological advancements and market niches; more importantly, they will also always be up-to-date and fully aware of their competitors’ technological directions. Engineers typically spend days or weeks studying patent documents one-by-one and aggregate the acquired knowledge manually; and this process goes on as there are new patents being filed each day. Indeed, the client has welcomed the prospect of text analytics and machine learning to automate this time consuming task, allowing them to put more focus into real R&D.

quandatics-mecha-infographic

Solution:
Coming from an engineering environment, the client lacked the tools and relevant expertise in advanced analytics, even though they knew what they wanted to achieve.  A document and text processing solution based on Quandatics’ advanced analytics platform was proposed and implemented for the client. The client provided domain specific inputs throughout the development process, as the understanding of the data – technical keywords and types of patent – was vital for building a classification model.

Meanwhile, Quandatics applied its know-how on advanced analytics to achieve the client’s objectives. By applying different text processing techniques in the analytics process, a variety of tasks – information extraction, transformation, filtering, word occurrence and frequency counting – can all be performed within a short period of time. What’s even better is the user-friendliness of Quandatics’ solution allowed the client to amend, enhance, or even create new patent analysis applications to suit their on-demand requirements, depending on the objectives of their investigations.  Finally, the output of the patent analysis process can be exported to a reporting environment, in this case Excel, to be visualized on graphs which inform the users about patents found that meet their search criteria.

quandatics-mecha

Advanced text analytics on international patents using Quandatic’s analytics platform

Business Benefits:
The previously highly manual and time-consuming patent analysis process is shortened by more than 15-fold, enabling the extra hours saved to be utilized to conduct core R&D activities. The self-service platform makes the modification of process possible and fully controllable by any users, reducing overhead costs and dependency on vendors or IT departments.

The Future:
More advanced clustering and learning methods can be applied on the structured patent data and more accurate, multi-level classification can be made available.  All the developed processes are reusable and have the potential to be replicated across other business units in the client’s organization, both locally and overseas, reducing the overall cost of performing patent analysis.

, , , , , ,