Beijing Humanoid Robot Half-Marathon Hits Final Test

// 45d agoVIDEO

Beijing Humanoid Robot Half-Marathon Hits Final Test

Beijing’s humanoid robot half-marathon just completed its full-scale rehearsal ahead of the April 19 race, with more than 70 teams testing autonomous and remote-controlled systems on the real course. The clip shows how robot sports are becoming serious embodied-AI stress tests for balance, battery life, gait control, and navigation.

// ANALYSIS

The spectacle matters less than the measurement: these races are turning humanoid robotics into a public benchmark for autonomy, not just a demo reel.

–The 2026 test covered the full 21.0975 km route and simulated timing, logistics, and emergency response.
–Nearly 40% of teams are targeting autonomous navigation, which is the real differentiator versus teleop.
–Long-distance running exposes the hardest robotics problems: energy management, dynamic balance, and stable gait under changing conditions.
–Public competitions like this generate real-world data that lab videos can’t, especially for urban mobility and recovery from missteps.

// TAGS

roboticsbenchmarkbeijing-humanoid-robot-half-marathon

DISCOVERED

45d ago

2026-04-17

PUBLISHED

45d ago

2026-04-17

RELEVANCE

7/ 10

AUTHOR

Anen-o-me

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

UPDATE31m ago

Ultracode demands stacked Claude Max subscriptions

A developer has evaluated the practical cost of running Claude Code's high-effort "Ultracode" mode, stating that the feature requires at least two Claude Max subscriptions to be viable for professional workflows. Although they spend $600 monthly on three Claude Max subscriptions to sustain their agentic coding tasks, they consider it the best money they spend due to the substantial productivity benefits.

LAUNCH38m ago

Google launches new agent developer tools for the Gemini Enterprise Agent Platform at I/O 2026, featuring Antigravity 2.0, Managed Agents API, and Agent Studio to streamline secure, production-ready AI agent deployments.

At Google I/O 2026, Google introduced a powerful suite of developer tools for the Gemini Enterprise Agent Platform to simplify the process of building, testing, and deploying autonomous enterprise AI agents. The new suite accommodates both low-code and code-first workflows: Agent Studio provides a visual workspace for non-technical teams, while the Agent Development Kit (ADK) 2.0 and Antigravity 2.0 (formerly the Gemini CLI) offer code-first developer environments for complex, multi-agent systems. Crucially, the new Managed Agents API allows developers to launch fully sandboxed, secure, Google-hosted agents with a single API call, resolving key infrastructure and security issues for production deployments. Integrated with Gemini 3.5 Flash and the open Model Context Protocol (MCP), these tools bridge the gap between local prototyping and secure cloud deployment.

UPDATE39m ago

Lovable projects ship with TanStack Start

AI full-stack application builder Lovable now builds and ships all new projects using TanStack Start as the default meta-framework. This integration provides developers with full-stack React capabilities out-of-the-box, enabling flexible per-route routing options (SSR, SSG, or CSR) alongside built-in support for server functions.

Beijing Humanoid Robot Half-Marathon Hits Final Test