Campus as an AI Sandbox: Open-Source Smart Building Infrastructure for Agentic AI Education

Wednesday, September 30, 2026 | 9:30AM–10:30AM MT
Session Type: Poster Session
Delivery Format: Poster Session
What if campus infrastructure—lighting, HVAC, elevators, and energy systems—could become a fully open, student-modifiable AI sandbox? This poster introduces an open-source, five-layer smart building platform spanning a data ingestion layer; a vendor-agnostic device abstraction layer; a unified API Gateway layer; a custom control logic layer; and an AI agent layer deployed across multiple campus buildings. We will discuss how students design and deploy real agentic AI applications through project-based learning, presenting a concrete framework for transforming closed campus infrastructure into a transparent, extensible Living Laboratory for responsible AI education. Attendees will obtain practical insights and an architectural model they can apply at their own institutions.

Presenters

  • Tomonori HASHIYAMA

    Professor, The University of Electro-Communications
  • Takuto Matsuhashi

    Assistant Professor, The University of Electro-Communications