# Robot Learning & Embodied AI: State of the Field (October 2026)

Research dossier · v1 · updated 2026-10-30 · 6,921 words · 111 sources · 28 claims · verification: verified

An expert briefing on robot learning as of late October 2026: vision-language-action (VLA) models, large-scale imitation learning, real-world reinforcement learning, sim-to-real transfer for legged and dexterous control, humanoid whole-body control, world models, evaluation and safety. The main takeaway is that generalist VLA policies, built on pretrained vision-language backbones with continuous action decoders and increasingly post-trained with reinforcement learning on real robots, now perform long-horizon household and dexterous tasks in settings not seen in training. The evidence base has not kept pace: the most-used simulated benchmark is saturated, real-robot trials are small and rarely blind, many 2026 claims are preprints or company announcements, and no certification route yet exists for learned controllers on balancing humanoids.

## Contents

- Executive summary
- Background and core concepts
- Timeline of key milestones
- State of the art
  - VLA model families
  - Data: teleoperation, human video and synthesis
  - Scaling and transfer
  - Reinforcement learning on real robots
  - Locomotion and humanoid whole-body control
  - Dexterous manipulation
  - World models and world-action models
  - Evaluation practice
  - Latency, safety layers and certification
  - Quantitative benchmarks
- Open problems and active debates
- Major programs and players
- Outlook: what to watch (November 2026 to April 2028)
- Key papers
- Glossary

## Sample findings

- RT-2 co-fine-tuned a web-pretrained vision-language model on robot trajectories, writing actions as text tokens, and was tested in about 6,000 trials.
- The open 7-billion-parameter OpenVLA, trained on 970,000 demonstrations, beat the 55-billion-parameter RT-2-X by 16.5 points of absolute success over 29 tasks.

## Get the full dossier

GET https://space-kitty.com/api/v1/dossiers/robotics-embodied costs $0.20 via x402 (USDC on Base Sepolia (testnet)). Send `Accept: text/markdown` for Markdown or `application/json` for structured data (claims with citations, timeline, benchmarks, open problems, glossary, bibliography).
---
Space Kitty Archive · https://space-kitty.com · API: https://space-kitty.com/api/v1 · MCP: https://space-kitty.com/mcp · Premium: $0.20/request via x402 (Base Sepolia (testnet)). Metadata: arXiv (CC0), OpenAlex (CC0), Europe PMC (CC BY/CC0 works), ClinicalTrials.gov.