Riemann Dynamics has partnered with Guanglun Intelligence and Noitom Robotics to build a comprehensive dataset for embodied AI. The collaboration aims to collect one million hours of embodied data by 2026, focusing on high-quality data acquisition, model training, and real-world robot deployment. This initiative will enhance Riemann Dynamics' Riemann-1.0 embodied world action model and Matrix-Game 3.5 interactive world model by integrating human behavior, force feedback, and environmental interaction data. AI
IMPACT This collaboration aims to accelerate the development of general-purpose embodied AI models by creating a massive, high-quality dataset, potentially leading to more capable robots in real-world applications.
RANK_REASON Partnership announcement focused on building a large-scale dataset for embodied AI, aiming for a specific future milestone. [lever_c_demoted from significant: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
