On April 19, the 2026 Beijing Yizhuang Half Marathon & Humanoid Robot Half Marathon (hereafter referred to as the “Robot Marathon”) officially kicked off. One of the biggest highlights of this year’s event was the large-scale participation of humanoid robots running fully autonomously.

Qianxun provided a unified spatiotemporal reference system for the event and delivered its BeiDou-powered Spatiotemporal Intelligence Trinity Kit to more than two-thirds of the autonomous navigation teams, enabling centimeter-level positioning and allowing robots to operate without remote control.

 

 

From Remote-Controlled to Autonomous: A Major Leap in One Year

This real-world competition served as a large-scale testbed, showcasing rapid advancements in humanoid robotics over the past year:

• Completion time improved significantly, with robots approaching the 1-hour half marathon mark
• Autonomous navigation teams increased to ~40%, compared to predominantly remote-controlled robots last year
• Capabilities evolved from simple demonstrations to complex environmental adaptability

Behind this progress lies continuous innovation in spatiotemporal intelligence, artificial intelligence, and motion control, forming the technological foundation that is bringing robots from laboratories into real-world environments.

The event was co-hosted by the Beijing Municipal Government, China Media Group, the Chinese Institute of Electronics, and the World Robot Cooperation Organization, with Qianxun providing timing and positioning support.

 

 

Over Two-Thirds of Autonomous Teams Powered by Qianxun

In the 2025 event, most robots relied heavily on manual remote control. In contrast, this year saw over 100 participating teams, with approximately 40% adopting autonomous navigation—making it the largest humanoid robot half marathon for autonomous operation to date.

How Do Robots Run Autonomously Over 21 km?

According to Wu Liang, Head of Embodied Intelligence Strategic Partnerships at Qianxun:

“Robots must independently complete positioning, path planning, motion control, and real-time decision-making across a 21.0975 km open urban environment.”

Accurate positioning is critical:

• Any deviation may cause route errors
• Or even collisions with barriers in complex environments

To address this, more than two-thirds of autonomous teams adopted Qianxun’s Spatiotemporal Intelligence Trinity Kit, which provides:

• Satellite + Ground integrated positioning
• Cloud + Chip collaborative computing
• Software + Hardware full-stack solution

This enables:

• Centimeter-level positioning
• Stable orientation and attitude sensing
• Reliable execution of navigation and motion control

 

 

Case Study: “Linglong” Robot Demonstrates Advanced Stability

At the event, the “Linglong” robot from the National and Local Co-built Humanoid Robotics Innovation Center showcased exceptional dynamic balance and full-body coordination.

The robot integrates Qianxun’s lightweight Arc-Unit positioning terminal, designed specifically for robotic navigation.

Key capabilities include:

• Proprietary “Lighthouse” positioning & orientation algorithm
• 0.1° orientation accuracy with a 1-meter baseline
• Continuous centimeter-level positioning output
• Robust performance under:
  • Signal interference
  • High vibration
  • GNSS signal loss scenarios

This ensures reliable autonomous navigation even in challenging race conditions.

 

Unified Spatiotemporal Benchmark for Fair and Transparent Competition

Beyond robot navigation, Qianxun also provided a unified spatiotemporal infrastructure for the entire event.

All 300+ participating robots were equipped with Qianxun’s BeiDou-powered smart shoulder badge, which:

• Transmits real-time data including:
  • Position
  • Speed
  • Distance
  • Time
• Integrates directly into the race timing system

This enables:

• Accurate ranking and timing
• Real-time monitoring of each robot
• Post-race performance analysis

During the event, Qianxun ensured 7×24 continuous service availability through:

• Dynamic cloud-based network optimization
• High-performance computing resource prioritization
• Real-time positioning performance monitoring

 

 

Spatiotemporal Intelligence: The Foundation for Real-World Robotics

“Every large-scale industry depends on foundational infrastructure,” said Dr. Chen Jinpei, CEO of Qianxun.

Over the past decade, Qianxun has been building an invisible “spatiotemporal infrastructure layer” connecting the digital and physical worlds.

Spatiotemporal intelligence enables:

• Translating digital commands into physical actions
• Providing robots with awareness of:
  • Where they are
  • When to act
  • How to move

 

Global High-Precision Positioning Network

As a global leader in spatiotemporal intelligence, Qianxun has established a worldwide network based on:

• BeiDou satellite system (compatible with GPS, GLONASS, Galileo)
• 10,000+ GNSS augmentation stations worldwide
• Proprietary algorithms, chips, and cloud platform

Delivering:

• Centimeter-level positioning
• Millimeter-level sensing
• Nanosecond-level timing

Applications include:

• Autonomous driving
• Low-altitude economy (drones, UAM)
• Humanoid robotics
• Smart cities
• Industrial digitalization

 

From Extreme Environments to Real-World Deployment

In early 2026, Qianxun successfully supported humanoid robots operating in -47°C conditions in Altay, Xinjiang, where robots:

• Completed 130,000+ autonomous steps
• Precisely traced the Olympic rings pattern on snow

This demonstrated robust performance in extreme environments.

 

What’s Next: Embodied Spatiotemporal Brain

Looking ahead, Qianxun is preparing to launch its next-generation product suite:

Embodied Spatiotemporal Brain

This system will:

• Integrate large foundation models with proprietary robotics models
• Enable:
  • Indoor & outdoor autonomous navigation
  • Environmental perception
  • Multi-robot coordination

It aims to accelerate the transition:

• From remote control → full autonomy
• From demonstration → real-world operations
• From single robots → swarm intelligence

Use cases include:

• Inspection
• Security
• Emergency response

 

 

Conclusion

With continuous advancements in spatiotemporal intelligence, industries such as:

• Humanoid robotics
• Autonomous driving
• Low-altitude economy

are rapidly moving from experimental stages to real-world large-scale deployment.

Qianxun is positioning itself at the core of this transformation—providing the critical infrastructure that enables intelligent machines to operate reliably in the physical world.