AI agents in AEC: how a remote road survey automated its own file workflows

​A road construction project in a remote Greek village with no fibre, no fixed infrastructure and challenging access. Exactly where typical AEC file workflows fall apart.

​But this isn’t your typical AEC workflow.

​Join LucidLink and Diorama.Studio to see how the studio captures live site data using drones and laser scanners, streaming it directly into the design office while the survey is still running, using only the connectivity available on site.

​Then the part nobody else is doing yet.

​Hermes, Diorama’s AI agent, connects to a live LucidLink filespace through the LucidLink MCP Server (beta) and takes over the post-capture work. It pulls imagery from every drone flight, renames files according to the studio's SOP and files them into processing-ready folders.

​Every battery swap used to reset the drone's file numbering, costing hours of manual renaming. Now no one touches a filename.

​What you'll learn:

• ​Why the site-to-office gap, not storage size, is the real bottleneck on infrastructure projects

• ​How to stream drone and laser-scan data off a remote site while capture is still running

• ​How AI agents work on live project files over MCP: reading, renaming and filing to SOP, with file locking and an audit trail

• ​How to build the same SOP-driven automation for your own workflows, using the LucidLink MCP Server (beta)