Skip to content
Agientry
ZH
← Back to Videos

Compass 1: One Navigation Model Across Four Robots

Loading video
Loading video

Archon Robotics combines sparse keyframe memory and spatial reasoning so one model can navigate a quadruped, humanoid, wheeled robot and drone from monocular RGB video.

NavigationFoundation ModelsEmbodied AIMonocular RGBCross-EmbodimentArchon RoboticsNavigation Base Model
Category: research
Author: @Archon_Robotics
Date: 2026-11-12T00:00:00
Duration: 174.59s
Source: https://x.com/Archon_Robotics/status/2120824552167227833
Reference: https://www.archon.tech/research/compass1

Related Videos

A-0 Trains Robot Policies on 16,000 Hours of Human Video
0:39

A-0 Trains Robot Policies on 16,000 Hours of Human Video

@yh_kris · 168 views · 2026-11-04
Human demonstrationsRobot Learning
SceneBloom: Simulation-Ready Scenes for Embodied Learning
0:38

SceneBloom: Simulation-Ready Scenes for Embodied Learning

@l_yuhan7272 · 178 views · 2026-10-28
SceneBloomEmbodied AI
Odyssey-3: A Foundation World Model for Physical Agents
1:37

Odyssey-3: A Foundation World Model for Physical Agents

@odysseyml · 1.3k views · 2026-09-15
World ModelsFoundation Models
AGC-VLN: Air-Ground VLN via Shared Bird's-Eye Maps
1:59

AGC-VLN: Air-Ground VLN via Shared Bird's-Eye Maps

@heetezition · 637 views · 2026-09-07
VLNAir-Ground Collaboration
BeingBeyond: Beijing embodied-AI startup building a pure robot brain
2:34

BeingBeyond: Beijing embodied-AI startup building a pure robot brain

@ErenChenAI · 544 views · 2026-08-08
BeingBeyondEmbodied AI
τ0-VLA: Robot VLA that searches over subtask sequences at test time with a world model
3:33

τ0-VLA: Robot VLA that searches over subtask sequences at test time with a world model

@jiqizhixin · 948 views · 2026-08-07
VLAWorld Models

As an Amazon Associate, we earn from qualifying purchases.