Artificial intelligence sports hardware company Pongbot has closed a Series B funding round valued in the triple-digit millions in Chinese renminbi (RMB). The financing round was led by the Shanghai State Investment Pilot Fund, with substantial participation from BlueRun Ventures and Pudong Venture Capital. The newly secured capital is designated to accelerate the development of specialized artificial intelligence systems and automated training hardware across athletic disciplines.
At the heart of Pongbot's technological portfolio are multimodal Sports Large Models designed specifically for physical athletics. Unlike general text or vision foundations, these domain-specific models focus directly on the rapid comprehension, modeling, and anticipation of athletic motion. Pongbot integrates these large-scale models with dedicated motion-vision sensors that capture physical movements directly on the court. This integration bridges algorithmic reasoning with responsive mechanical hardware in a unified operational loop.
Pongbot originally established its footprint in table tennis, designing autonomous training robots capable of rapid ball delivery and practice drills. Leveraging that technical base, the company has expanded its focus to tennis courts by launching advanced product lines. Central to this expansion are the PACE series and the specialized training robot named Aura. These machines represent an evolution away from traditional stationary ball launchers toward responsive, artificial-intelligence-driven training partners.
The operational foundation of these new robots depends on continuous real-time data capture. During active rallies, the integrated motion-vision sensor array tracks stroke dynamics and assesses the biomechanics of the practicing athlete. Simultaneously, the onboard vision systems calculate the precise flight trajectories and velocity of the ball. This continuous measurement loop tracks how cleanly a stroke is executed and evaluates physical movement patterns without introducing operational latency.
Based on this continuous data evaluation, Pongbot's robotic systems autonomously alter their operational parameters. The machines dynamically adjust ball placement across the court, modify the type and magnitude of spin, and vary the rhythm of successive shots. This real-time adaptability operates without requiring manual intervention from human coaching staff standing at the sideline. The system effectively counters the player's returns by presenting evolving tactical challenges.
This autonomous flexibility enables Pongbot's platform to serve diverse user tiers within the athletic spectrum. The models can be configured to guide young prospects through repetitive technical drills and fundamental stroke development. At the same time, the hardware can simulate the demanding cadence and spin variations required by elite professional athletes. Through this dual focus on biomechanics and robotics, the technology demonstrates the expanding utility of specialized AI models in modern training environments.