cd /news/robotics/figure-ais-helix-2-5-solves-the-gene… · home › topics › robotics › article
[ARTICLE · art-133754] src=asiaai.fyi ↗ pub= topic=robotics verified=true sentiment=↑ positive

Figure AI’s Helix 2.5 Solves the Generalization Problem in 30 Unknown Homes

Figure AI unveiled Helix 2.5, a new neural network for humanoid robots, on September 17, 2026, demonstrating robots performing extended household chores such as tidying, towel folding, and bed making in 30 homes where no prior data collection had occurred. The company said Helix 2.5 uses a pre-trained foundational model to generalize across varied, unknown settings, unlike its earlier Helix 02, which required extensive data collection within the specific operational environment. Figure AI framed the result as a step toward practical general-purpose humanoids that move beyond environment-specific learning.

by read1 min views1 publishedSep 18, 2026
Figure AI’s Helix 2.5 Solves the Generalization Problem in 30 Unknown Homes
Image: Asiaai (auto-discovered)

Figure AI’s Helix 2.5 Solves the Generalization Problem in 30 Unknown Homes

Figure AI unveiled Helix 2.5, a new neural network for humanoid robots, on September 17, 2026.

AsiaAI Publisher · September 18, 2026 · 2 min read · Source: ロボスタ Robot Start · Issue #100

East Asian Technology Intelligence

Japan & China tech news — translated, contextualized, and delivered for Western readers.

Free. Unsubscribe anytime.

This story ran in Issue #100, alongside three other stories.

Robotics & Automation

Figure AI unveiled Helix 2.5, a new neural network for humanoid robots, on September 17, 2026. The company demonstrated its robots performing extended household chores like tidying, towel folding, and bed making in 30 homes where no prior data collection had occurred, showcasing adaptation to unknown environments. This represents a step towards practical general-purpose humanoids, moving beyond environment-specific learning.

Previous humanoid robot deployments, including Figure AI’s own Helix 02, required extensive data collection within the specific operational environment. Helix 2.5 aims to overcome this limitation by using a pre-trained foundational model and demonstrating generalizability across varied, unknown settings.

source & further reading
asiaai.fyi — original article Huawei’s Next-Gen Ascend: Why Advanced Interconnects Are Key to Challenging Nvidia’s Clusters Anthropic’s 30,000-Agent R&D Engine: Claude Code Shifts 26% of Core Engineering to Autonomy OpenAI’s Misalignment Framework: A Tactical Bid to Preempt Global AI Governance
── more in #robotics 4 stories · sorted by recency
── more on @figure ai 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
→ Live at https://your-agent.zahid.host ✓
Get free account → Pricing
from €0/mo · no card required
LIVE [news/figure-ais-helix-2-5…] indexed:0 read:1min 2026-09-18 · —