Can robots learn to reason, recover and act more naturally? An embodied AI platform aims to bridge that gap with unified intelligence.
ACE ROBOTICS has unveiled Kairos 3.1, an action-oriented world model designed to improve how robots perceive, reason and interact with the physical world. Alongside the model, the company introduced Ambient Capture Engine 2.0, three commercial embodied AI solutions and a new industry benchmark, expanding its technology stack from data acquisition to real-world deployment.
The launch addresses a key challenge in embodied AI: enabling robots to move beyond perception and execute tasks reliably in dynamic environments. Rather than relying on separate AI models for different functions, Kairos 3.1 combines generative, physical and cognitive intelligence within a single architecture. The model processes visual inputs, language instructions, force and tactile feedback, and motion trajectories in a unified latent space, allowing robots to understand tasks, plan actions, execute them and evaluate the outcomes.
According to the company, the platform can break down long-horizon tasks into smaller steps, simulate multiple action paths before execution and adapt its strategy if an action fails instead of restarting the entire process. During internal testing, the 8B model achieved an inference latency of 125 ms on NVIDIA’s Jetson Thor platform, supported by the in-house KairosRT engine for on-device, real-time inference.
Complementing the model is Ambient Capture Engine 2.0, a human-centric data collection system designed to capture high-density physical interaction data through wearable sensors, force-sensitive gloves and automated annotation tools. The company also introduced ACE-Data-0, an open household interaction dataset, and PHYSICAL IQ, a benchmark developed with research institutions and industry partners to evaluate embodied intelligence systems.
“Kairos 3.1 is built around a first-principles approach to embodied world models, helping robots act more reliably in complex and uncertain environments, and accelerating the arrival of physical AI’s Kairos moment,” says Wang Xiaogang, Chairman of ACE ROBOTICS.
Click here for the official announcement.
