On September 7, 2026, Polaris.AI announced that it had begun offering its “Design × AI Utilization Solutions” for the manufacturing sector. The solution targets design and development work in fields including machinery and equipment, electrical and electronic systems, automobiles, and others, using artificial intelligence to handle engineering drawings, design data, and related workflows.
The company says the solution is not limited to creating a new drawing, but links drawings with engineering standards, knowledge records, and manufacturing data. It covers five main categories of design and development work: searching previous drawings, supporting design using retrieval-augmented generation (RAG) technologies, creating and editing drawings, reviewing designs, and supporting production-process design.
What does the solution offer?
The offering includes functions for handling a drawing from the time it is created through its management after approval. The capabilities mentioned include support for creating new drawings, analyzing drawing structures, assisting designers in verifying elements, and managing drawings, specifications, and changes. It also enables the comparison of drawings and specifications, tracking the history of modifications, and analyzing the scope of a change’s impact.
On the manufacturing side, the system can be used to extract dimensions and components from drawings, then output information such as specifications or catalog data. It also aims to support process design and link information contained in the drawing with inspection and measurement data and operational comments, culminating in a final review that helps reduce omissions before the design is delivered.
Three technical foundations
Polaris.AI builds the solution on three axes: understanding drawing structures, converting design standards into usable knowledge, and linking design with manufacturing and operations. In structural analysis, drawings or images of them and data on target shapes can be converted into a graphical representation of the relationships among elements, allowing vision and language models to understand the relationships within a drawing rather than treating it merely as a separate image.
The standards axis is based on the customer’s design standards, tolerance rules, and expert notes, with the aim of producing interpretable outputs according to engineering logic that can be reviewed. The company says the process-linkage axis integrates inspection and measurement results and comments into the design workflow, allowing previous information to be used when making subsequent decisions.
What changes in practice?
If these functions succeed in real-world environments, the core value lies not merely in automatically generating a drawing, but in reducing the time required to find, understand, and compare previous drawings, and in transforming knowledge scattered across CAD and PDF files, catalogs, and specification data into a reference that can be queried. Polaris.AI states that verification also relies on using previous projects and accumulated verification results to reduce uncertainty in design decisions.
The company also indicates that it is conducting research to interpret relationships in two-dimensional drawings and reconstruct three-dimensional CAD models from them. However, the material does not specify the solution’s prices, availability mechanism, customer names, or independent quantitative results measuring accuracy and productivity. Therefore, the practical benefit remains tied to the quality of the drawing data and standards provided by each manufacturer, as well as engineering teams’ ability to review the outputs of the artificial intelligence before approving them.