Researchers at Carnegie Mellon University have developed a small device called "Object-Agent Twins" (OATs) that attaches to everyday objects and displays their status through character-like expressions. These devices, costing under $20 to build, utilize an ESP32-S3 board, a monochrome e-paper display, and various sensors. An AI model, Google's Gemini, generates character animations based on user-defined object characteristics and emotional models, with the system designed to conserve battery life by communicating with a server intermittently. Initial user testing showed that OATs can increase user engagement with objects and provide insights into their condition, though the researchers acknowledge potential for anthropomorphism and plan to enable on-device processing for future iterations. AI
IMPACT Could lead to more intuitive human-object interaction and novel forms of ambient computing.
RANK_REASON Research paper announcement from a university lab. [lever_c_demoted from research: ic=1 ai=1.0]
Read on Mastodon — mastodon.social →
- Carnegie Mellon University
- Chris Harrison
- Daehwa Kim
- Detroit
- ESP32-S3
- Gemini
- GitHub
- Object-Agent Twins
- UIST 2026
- Waveshare
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →