Problem & Context:

The client was interested in learning about materials that are used in valves, especially aerosol valves, which are made of plastics. The client wanted SPA to explore and prepare a report on type of valve materials that were being studied globally.

SciTech Patent Art Approach & Deliverables:
  • The SPA team conducted a patent search and identified more than 5000 relevant records.
  • SPA developed a ML/AI tool to analyze 5000 records would have been a very time intensive task.
  • A 3-member SPA team trained the AI/ML tool.
  • Subsequently, SPA team used this trained model to categorize a sample patent data-set. This data-set formed the core for further training of the AI/ML tool.
  • The above Step was iterated several times to improve the accuracy of the output.
  • The AI/ML model was thus made ready following which the entire 5000 patent data-set was uploaded on the tool and was used for categorizing the records based on type of valve, material of construction etc.
  • The technology assessment was thus made based on a huge patent data-set using an in-house developed AI/ML tool.
  • The model achieved more than 70% accuracy.
  • The AI/ML tool was integrated into a globally accessible portal which enabled quick visualization of the patent data-set using graphical representation.
Turnaround Time:
  • 2 weeks
  • $ 25,000