Ropedia launches academic partner program for Physical AI research

Ropedia said it is opening access to its HOMIE Gen2 capture device for selected academic researchers and institutions as it expands its Physical AI data business. The Singapore startup also launched a 500-hour evaluation program and plans to appear at IROS 2026 in Pittsburgh.

Why it matters: - Physical AI and embodied AI research depends on real-world multimodal data, not just lab settings. - Ropedia is trying to make that data easier for universities and research labs to capture, process and use for model development. - The move could widen access to infrastructure for work on robotics, world models and vision-language-action models.

What happened: - Ropedia launched an Academic Partner Program for selected academic researchers and institutions. - The program gives participants access to HOMIE Gen2, Ropedia’s head-mounted multimodal capture device. - The launch is aimed at advancing research in robotics, embodied intelligence and human-robot interaction. - Ropedia also launched a 500-hour Ropedia Evaluation Program for real-world Physical AI data. - The company said it is increasing its focus on the U.S. market and will attend IROS 2026 in Pittsburgh from September 28-30 at booth 742. - Ropedia is headquartered in Singapore and has an office in Mountain View, California.

The details: - HOMIE Gen2 captures synchronized multimodal information from real-world environments, including visual, motion, spatial, inertial and semantic data. - Researchers can use their own processing pipelines or Ropedia’s data processing pipeline. - The resulting data can support development of World Models and Vision-Language-Action models. - Ropedia’s Xperience-10M dataset has been downloaded more than 2.3 million times. - The dataset supports research in egocentric understanding and embodied AI. - Ropedia’s academic collaboration network includes Princeton University, Carnegie Mellon University, the University of California San Diego, the University of Illinois Urbana-Champaign, The Hong Kong Polytechnic University, Westlake University, Allen Institute for AI and Advanced Machine Intelligence. - Researchers and institutions can apply for a HOMIE Gen2 Research Kit at ropedia.com/collaboration. - Ropedia says its broader data infrastructure covers real-world capture, data processing, annotation and model-ready datasets. - The company says Xperience-10M is open source. - Ropedia says it serves more than 20 global robotics and foundation model companies.

Between the lines: - The program is a way to move Ropedia deeper into the academic research pipeline while reinforcing its platform as infrastructure, not just a data source. - The open-source dataset strategy can help build adoption and keep Ropedia’s tools relevant as Physical AI research becomes more competitive. - The U.S. push suggests Ropedia is positioning for more direct engagement with robotics and AI labs in a key market.

What's next: - Academic teams can apply now for participation in the partner program. - Ropedia is expected to use its IROS presence to reach researchers and potential customers in robotics. - The 500-hour evaluation program adds another route for groups to test Physical AI workflows with real-world data.

The bottom line: - Ropedia is broadening access to its Physical AI data stack as it tries to become a core infrastructure provider for robotics and embodied AI research.