Librizer AI circuit library creation tool adds KiCad support, reducing design man-hours by automatically generating symbols and footprints from datasheets

Leacron, a startup originating from Tokyo University of Science (located in Shibuya, Tokyo; CEO: Yuta Takayasu), has announced that its design support tool, 'Librizer,' which uses AI to automatically generate schematic symbols and PCB footprints from electronic component PDF datasheets, now supports KiCad format.

With this update, designers using KiCad can generate schematic symbols and PCB footprints simply by uploading a PDF datasheet and utilize them as parts libraries within KiCad.

Librizer is a software tool aimed at automating the creation of parts libraries needed in electronic circuit design, thereby reducing the workload on designers.

■ Background of KiCad Support

In electronic circuit design, when using new parts, designers often need to organize pin numbers, names, package information, and dimensions from datasheets, and manually create schematic symbols and PCB footprints.

While many EDA tools come with existing parts libraries, new or specialized components often lack the required libraries. In such cases, designers have to manually create the symbols and footprints while referencing datasheets.

Particularly in R&D and prototyping environments, the frequent need to use new components can make library creation a bottleneck in the design process. At Leacron, the team felt the time-consuming nature of this task while pursuing their own hardware development. They developed Librizer with the belief that automating this process would allow designers to devote more time to essential circuit design and verification.

KiCad is an EDA tool widely used by individual developers, researchers, startups, educational institutions, and open-source hardware developers. This support makes it easier for a broader range of designers to integrate Librizer into their design workflows.

■ Key Features of Librizer

1. Automatic Schematic Symbol Generation
Librizer uses AI to analyze pin information in datasheets and automatically generates schematic symbols, including pin numbers, names, and configurations, making them easy to use in EDA tools. This significantly reduces manual entry and symbol creation time.

2. Automatic PCB Footprint Generation
Librizer generates footprints for PCB design based on package information in datasheets. By analyzing pad configurations, pin pitch, and dimensions, it streamlines pre-PCB design steps, reducing the need for manual dimension entry.

3. Batch Generation of Symbols and Footprints
Librizer can generate both schematic symbols and PCB footprints in one go. This allows for quick creation of usable parts libraries for EDA tools, shortening setup time and allowing designers to concentrate on the actual circuit design.

■ Proprietary Technology: 'Hybrid Approach' to Compensate for AI Weaknesses

To achieve high-precision PCB footprint generation, Librizer employs a hybrid approach combining AI analysis with classical geometric processing. The process involves three steps:
1. Acquisition of dimension information via VLM (Vision Language Model).
2. Estimation of geometric shapes using classical algorithms.
3. Re-acquisition of dimension information via VLM based on estimated shapes.

While VLMs alone can have errors when reading drawings or dimensions, combining them with classical geometric analysis significantly improves precision. This same technology is applied to the new KiCad support to ensure the output is ready for KiCad design workflows.