BiMatrix Completes Year 1 Development of AI-Powered Integrated Permit Pre-Diagnosis Service
IT DAILY ·

✦ AI Summary
As the lead organization for the development and demonstration project for an "AI-based integrated permit pre-diagnosis service," BiMatrix announced that it had completed the key Year 1 development work.
The project is included in the Ministry of Science and ICT's "Top 10 AI Livelihood Projects," with the National Information Society Agency (NIPA) as the operator and the Ministry of Land, Infrastructure and Transport as the demand agency.
The service is designed to integrate land and parcel permit laws and ordinances from multiple agencies and use AI to pre-diagnose whether permits are possible before development activities begin, with field validation for pilot operation set to start in January 2027.
BizCrushWant to follow Korean talks and live streams in real time?Try CapCrush on the web
BiMatrix, the lead organization for the development and demonstration project for an AI-powered integrated permit pre-diagnosis service, announced on the 19th that it had completed the key Year 1 development work. The project is included in the Ministry of Science and ICT's "Top 10 AI Livelihood Projects."
The service aims to integrate land- and parcel-permit laws and ordinances scattered across multiple agencies, and use AI to pre-diagnose whether permits are possible before development activities begin. To that end, ontology, retrieval-augmented generation (RAG), multi-AI agents, and 3D simulation technologies are being applied.
The project operator is the National Information Society Agency (NIPA), and the demand agency is the Ministry of Land, Infrastructure and Transport. The project is being carried out under a structure led by BiMatrix, with Wavers and ICT Way participating in the consortium.
After completing the key Year 1 development work, BiMatrix will move on to field validation for pilot operation in January 2027 as the next step. The project period runs through December 2027.
While carrying out the Year 1 project, the BiMatrix consortium built an integrated permit database (DB) and a knowledge base, implemented multi-AI agents, and developed the pre-diagnosis platform. Through this, it completed the service foundation. On the 4th, a performance report session hosted by the Ministry of Land, Infrastructure and Transport and the Ministry of Science and ICT was held, and the development results were shared there.
The consortium plans to obtain performance certification from an external specialized institution by the end of the year. It will then conduct field validation with officials from 5 pilot-operation local governments and local permit agents, and is expected to carry out final user testing. Through this process, it aims to complete preparation for pilot operation.
This effort is based on the fact that the existing land development permit preparation process consists of 6 steps. The 6 steps are land search, regulatory analysis, expert consultation, land selection, budget formulation, and permit application. In addition, the complexity of laws and ordinances and differences in detailed standards by local government created a high dependence on experts. In many cases, permits were denied even after time and money had already been spent.
A new pre-diagnosis service has been developed. It is designed to immediately guide users on whether permits are possible, the complaint-handling process, and the required documents based on the land location and intended use entered in natural language by applicants.
The service also supports 2D map-space analysis and 3D simulation functions. With 3D simulation, users can check building placement, access to sunlight, and views. As a result, the permit preparation process is expected to be reduced from 6 steps to 3, and the burden of initial conceptual design costs amounting to several million won is also expected to ease.
BiMatrix is the overall lead company. It was responsible for building the ontology knowledge base based on relationships among zoning areas, laws, and regulatory restrictions; developing the pre-diagnosis AI inference model; implementing multi-AI agents; implementing law-knowledge search (RAG); and applying and demonstrating the service in the field. The consortium reflected the characteristics of permit work, where local ordinances differ by local government and legal grounds must be presented. It also structured about 135,000 articles and clauses from laws and ordinances, along with 190 GB of spatial data, into an ontology knowledge graph so that the legal basis for AI judgment results can be traced back.
The multi-AI agent consists of an integrated orchestrator, a pre-diagnosis inference agent, and a spatial analysis integration agent, and was implemented so that each handles its own area. The photo shows the analysis results of the "AI-based integrated permit pre-diagnosis service," courtesy of BiMatrix.
The service plans to begin pilot operation first in Gyeyang District in Incheon, Eumseong County in North Chungcheong Province, Asan City in South Chungcheong Province, Gyeongsan City in North Gyeongsang Province, and Yeongcheon City in North Gyeongsang Province in January 2027, before expanding its coverage to 18 local governments in the future.
In the next year, the number of permit-diagnosis targets will expand from 10 to 19. The second-year additions will include conditional alternative proposals that guide users on how to meet permit requirements, optimal site recommendations, mobile services, and administrative support functions for civil servants.
Jo Geun-sang, head of BiMatrix's AI permitting project team, said the Year 1 development confirmed that AI can understand and judge complex laws and ordinances that vary by region, together with the grounds for those judgments. He said the company plans to prepare for pilot operation without setbacks, while also aiming to connect the achievements of AI technological innovation to improved quality of life for the public. A BiMatrix official said that rule- and standard-based judgment work applies not only to permits but also broadly across public institutions and companies, adding that the technologies validated in this project will be provided to help more organizations make faster and more accurate decisions.
Source: IT DAILY · Kwon Young-seok
Original: https://www.itdaily.kr/news/articleView.html?idxno=243048
References
This article was produced with the help of an automated content generation algorithm.
Source: IT DAILY
View originalThis article was summarized and organized by BizCrush based on the original article from IT DAILY. For exact quotations and full details, please refer to the original article.
