Shengshu Technology Partners with Lightwheel AI to Bring World Models into the Real World

On June 15, Beijing Shengshu Technology Co., Ltd. (“Shengshu Technology”) and Lightwheel AI (Beijing) Technology Co., Ltd. (“Lightwheel AI”) announced a strategic partnership. The two parties will collaborate deeply on building data standards for general world models, establishing high-quality world model evaluation systems, and accelerating world model deployment in real-world scenarios, driving general world models toward being “physically credible, task-capable, and verifiable in real scenarios.”

Over the past two years, multimodal foundation models have made continuous breakthroughs in visual quality, generation duration, and motion consistency, transitioning from “content generation” to “world models” — becoming the core foundation for understanding world states, predicting futures, and planning actions. As world models move toward the physical world, embodied intelligence is one of the key applications. The core value of world models is no longer just generating content, but understanding physical laws, performing complex task reasoning, and supporting real-world actions through the interaction between digital and physical spaces.

This partnership represents a powerful alliance at this industry inflection point: Shengshu Technology leverages its deep expertise in general world models, centered on its World Action Model MotuBrain, to build a universal embodied intelligence brain that unifies “predicting the world” and “driving action.” Lightwheel AI builds data and evaluation infrastructure for physical AI around human data, simulation evaluation, and deployment feedback. The two parties will focus on three joint initiatives: high-quality world model data systems, reproducible world model evaluation frameworks, and real-world scenario verification and continuous iteration mechanisms.

The capability boundary of world models fundamentally depends on the physical coverage and task diversity of data. To meet the needs of general world model training and iteration, both parties will combine Shengshu Technology’s technical expertise and multi-source heterogeneous data absorption capabilities with Lightwheel AI’s data definitions, quality systems, and best practices in human video data and synthetic simulation data, jointly exploring data standards and construction methods for world models.

As world models enter the physical world, progress cannot be judged by subjective visual impressions alone. Whether models truly understand physical laws, complete specific tasks, predict action outcomes, and meet conditions for real-world deployment all require a unified, stable, and reproducible evaluation system. Based on Lightwheel AI’s industrial-grade simulation evaluation platform RoboFinals, both parties will jointly explore evaluation tasks, metrics, and data quality standards for world models and embodied intelligence models, building a continuous iteration loop from capability diagnosis to data supplementation to model optimization.

The ultimate value of world models must be validated in real tasks and industrial scenarios. Both parties will conduct joint exploration around embodied intelligence and physical AI application scenarios, covering data collection, task SOP conversion, simulation evaluation, real-machine verification, feedback assessment, and continuous iteration.

Shengshu Technology will leverage its core capabilities in spatial understanding, physical law modeling, and end-to-end action decision-making, while Lightwheel AI will provide simulation scenarios, evaluation platforms, real-machine verification, and scenario delivery capabilities.

The strategic partnership between Shengshu Technology and Lightwheel AI represents not only complementary capabilities between a world model company and a physical AI infrastructure company, but also a key direction toward physical AI: both parties will continuously explore how world models can learn, reason, and validate in real environments, truly transitioning from the digital world to the physical world.

Looking ahead, Shengshu Technology will build a complete chain from perception to action through unified model architectures, establishing a “understanding the world — predicting the world — acting in the world” technical closed loop, making general world models truly the bridge between the digital and physical worlds.