Digital Twin Patent Families From Ten Major Industrial Firms

Andreea Toiu's dataset contains 396 patent families from ten major industrial firms like Siemens and IBM, covering 2003–2023. It supports replication for a study on digital twin patent functions, with each family classified into one of three functional categories.

Use Cases

• Analyze the distribution of Descriptive (D), Predictive (P), and Prescriptive (Pr) functional categories across the 396 patent families.
• Compare patenting activity over the 2003–2023 time range among the ten listed industrial firms like Siemens and General Electric.
• Investigate the relationship between patent family priority dates and their assigned functional categories.
• Validate the hierarchical rule-based coding procedure using the reported inter-coder reliability metrics (κ = 0.609 and κ = 0.912).

Strengths

• 396 patent families provide a consolidated view of inventions across jurisdictions.
• Covers a 20-year time range (2003–2023) from ten specified major industrial firms.
• Functional classification validated with reported inter-coder reliability (κ = 0.912 post-adjudication).

Limitations

• Small scope of 396 records and 91.5 KB file size limits statistical power for detailed subgroup analysis.
• Focus on only ten pre-selected firms may not represent the broader digital twin patent landscape.
• Classification relies on a specific rule-based procedure combining IPC and keywords, which may not capture all nuances.

Provenance

Source

PatBase, with assignee names harmonized using the OECD HAN database (December 2023 version).

Collection Method

Retrieved and functionally classified via a hierarchical rule-based coding procedure for a scholarly study.

Time Range

2003–2023.

Freshness

Data last updated March 2026.

Geography

Patent families consolidated across jurisdictions; specific geographic coverage not stated.