In 2024, MLIT newly announced i-Construction 2.0, which aims to achieve at least a 30% reduction in labor requirements at construction sites by fiscal year 2040, and is promoting various initiatives to this end. The presentation will introduce the latest case studies across these initiatives.

This presentation explains the concept of maturing AI agents—the next evolution beyond generative AI and RAG—through "process learning" rather than prompt engineering. Focusing on Interface Training for business tool operations, specialized models, and evaluation design, it demonstrates the key considerations for achieving real-world deployment while avoiding proliferation and obsolescence, as well as the competitive advantages of the Agent Economy. Examples of agent-based social simulations will also be presented.

This presentation covers the background that led our company to take on the challenge of remote construction, case studies of safety improvement through remote construction conducted in 2017 and 2020, and case studies of labor-saving and productivity improvement through remote construction conducted in 2024 (which received the Minister of Land, Infrastructure, Transport and Tourism Award in the FY2025 Infrastructure DX Grand Prize). Additionally, we will introduce infrastructure DX initiatives achievable even by regional construction companies, as well as prospects for the future of remote construction.

MLIT is advancing the "New Technology Utilization Scheme in Public Works," which promotes the use of useful new technologies. Additionally, this year, the ministry has formulated the 6th MLIT Technology Basic Plan, which aims for a virtuous cycle through innovation (Innovation Ecosystem), demonstrating the importance of technology improvement and institutional review through actual field applications.
This presentation will introduce the new MLIT Technology Basic Plan along with the latest initiatives related to new technology utilization and technology development in public works.

As infrastructure aging progresses, it is important to systematically advance maintenance, management, and renewal to achieve sustainable infrastructure maintenance. On the other hand, local governments face challenges in terms of finances and organizational capacity, making the full-scale transition to preventive maintenance difficult. Against this backdrop, this presentation will introduce MLIT's policies and initiatives, including "Regional Infrastructure Group Regeneration Strategy Management" and the "Infrastructure Maintenance Grand Prize."

In this lecture, I will introduce the implementation of an infrastructure digital twin that applies a voxel-based spatial representation similar to Minecraft, an initiative that received the FY2025 Infrastructure DX Grand Prize (Minister of Land, Infrastructure, Transport and Tourism Award). By utilizing a game engine that enables intuitive operation regardless of technical expertise and integrating point cloud data, CIM, and open data, I will discuss the potential of on-site DX to enhance construction and maintenance processes while addressing workforce shortages.

In the construction industry, where improving labor productivity has been a longstanding challenge, the use of generative AI has become not just an 'option' but an 'urgent necessity.'
However, the reality on the ground is that generative AI adoption often stalls because specific use cases and implementation scenarios remain unclear.
This presentation goes beyond mere technical discussions of generative AI, focusing on practical applications—'where, in which situations, and how to leverage it'—bringing the technology down to the operational level.

This seminar introduces Quasi‑Zenith Satellite System “Michibiki,” explaining its current operations and future expansion toward an 11‑satellite constellation. It also highlights the system’s high‑precision positioning services and their value as social infrastructure, along with practical examples showing how they improve productivity and safety in fields such as construction and surveying.