Coverage of WAIC: Daxiao Unveils Kairos World Model, Human-Centric Data Collection 2.0, and Three Industry Solutions

Deep News
Jul 20

News from the 2026 World Artificial Intelligence Conference and the High-Level Meeting on Global AI Governance, held in Shanghai from July 17th to 20th.

During the event, the "Six Dragons" World Model Summit Forum was hosted by Daxiao Robotics. At the forum, Daxiao Robotics announced three major core achievements: the Kairos World Model 3.1 (Kairos 3.1), the Environment-Style Data Collection Solution 2.0, and a suite of solutions covering three major vertical scenarios.

Kairos 3.1 is an action-integrated world model that fuses generative intelligence, physical intelligence, and cognitive intelligence. It breaks through the technical barriers of understanding, generation, and prediction with a natively unified architecture, constructing a full-chain self-evolving intelligent loop of "understanding-deduction-execution-reflection."

Within this framework, generative intelligence builds internal representations of the physical world, physical intelligence masters physical causality and deduces parallel worlds, and cognitive intelligence supports complex logical reasoning and long-horizon task decomposition. The deep unification of these three enables reliable and controllable closed-loop control.

Based on this unified architecture, the model forms a complete self-evolution loop. It first decomposes long-horizon tasks and assesses the environment through world understanding, then conducts physical causal deduction and multi-strategy calculation in parallel spaces to filter and output optimal action predictions, simultaneously completing process analysis and result evaluation while feeding state information back to the model.

The ACE-BRAIN-0.5 foundation solidifies the spatial groundwork for the Kairos world model, having cumulatively secured twelve global state-of-the-art (SOTA) achievements. Its capabilities across all dimensions—spatial understanding, navigation decision-making, operational execution, and progress assessment—place it in the global first tier.

In real-world household scenarios like laundry, robots can accurately identify spatial relationships between components to achieve autonomous closed-loop operation.

Regarding physical generation and deduction, Kairos-HomeWorld achieves whole-house generation and full object interaction for the first time. It is built with 300,000 sets of Chinese residential floor plans, 5,000 whole-house simulation scenarios, and 8,700 3D assets covering six physical properties.

In terms of performance, the Kairos 3.1 8B model boasts a reasoning latency of only 125 milliseconds, achieving real-time edge-side inference while maintaining the high intelligence ceiling characteristic of world models.

Concurrently, Daxiao also released its Human-Centric Environment-Style Data Collection Solution 2.0.

Specifically, on the data collection front, the ACE Ego Kit, featuring a split, lightweight, wireless "head-hand-chest" trinity design, ensures the collection of high-fidelity, comprehensive, "all-inclusive" data.

Moving to the annotation phase, the fully automated production platform, ACE Data Engine, takes over to achieve "fine-grained" labeling. Daxiao has also officially open-sourced the ACE-Data-0 dataset, an L5-level complex task dataset for home environments, to the entire industry.

Furthermore, Daxiao Robotics has developed three standardized industry solutions designed for rapid replication and deep adaptation to real-world business processes. These are the "Xiaoman" solution for instant retail applications, the "Xiaoxin" solution for hotel service scenarios, and the "Xiaotu" solution for autonomous operation in open scenarios.

During the same period, the Physical IQ Unified Benchmark Platform for Embodied Physical Intelligence was officially launched. This platform is jointly initiated by the Shanghai Artificial Intelligence Industry Association, the Hetao Institute, and the East China Branch of the China Academy of Information and Communications Technology, with participation from over 20 universities and industrial organizations.

Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.

Most Discussed

  1. 1
     
     
     
     
  2. 2
     
     
     
     
  3. 3
     
     
     
     
  4. 4
     
     
     
     
  5. 5
     
     
     
     
  6. 6
     
     
     
     
  7. 7
     
     
     
     
  8. 8
     
     
     
     
  9. 9
     
     
     
     
  10. 10