The Context Layer for Physical AI

We turn messy field imagery from cameras, drones, and inspections into structured understanding of assets, sites, risk, and action: the foundation for AI systems that can understand, monitor, and eventually operate in the physical world.

Records
Every asset, every site, every risk
Context
Source and confidence on every judgment
Featured in
WIREDNikkei
Backed by
Institute of Science TokyoOIST Innovation未来共創イニシアティブ
What we build

A record of every asset, site, and risk

Field judgments live in an expert's eye and end up on paper: photos in folders, findings in PDFs. We turn them into structured, machine-readable records of your assets. What it is, where it is, what condition it's in, which rules apply, and how confident the judgment is.

01

Field imagery

Whatever your sites already record with is enough. No new capture hardware required.

  • Fixed cameras
  • Drones
  • Smartphones
  • Inspection footage
02

A structured record

Every image linked to the right asset, location, time, and history. Every judgment carries its source and its confidence.

ASSET RECORDYD-2026-0417
assetSupport bracket B-17
location34.6873, 135.5259 · Zone C
conditionCorrosion, grade 2 (monitor)
history3 records · last 2025-11
rulesInspection manual §4.2
sourceIMG_0427 · 2026-05-30
confidence0.93
03

Ready for what reads it

Dashboards and reports for your teams today. APIs for the software that comes next, and eventually for the autonomous systems that follow.

  • Dashboards
  • Reports
  • APIs
  • Autonomous systems

Grounded in three layers of context

A record is only trustworthy if it knows where it came from. Every judgment is anchored in context.

Inner Context/ yours
  • Drawings
  • SOPs
  • Domain rules
  • Inspection history
Public Context/ open
PLATEAU
国土交通省
気象庁
国土地理院
IEC
ISO
ESA Sentinel-1
J-STAGE
Yodo Context/ ours
  • PLATEAU-based Japan Digital Twin
  • Coverage & Confidence Layer
  • Failure-Mode Atlas

Selected work

Case Studies
Research

Team

Xiuxi Pan, PhD

Xiuxi Pan, PhD

Technical Lead

Published work in Computer Vision, Computational Imaging, Generative Model. Led the production and implementation of AI solutions for major Japanese and international enterprises. Profiled by WIRED and Nikkei as an early pioneer of AI in Japan.

Sho Osawa

Sho Osawa

Commercial Lead

Former strategy consultant at Strategy& (PwC) and Westpac Banking Group. Scaled a SEA start-up insurer from zero to 150M in insured exposure in two years, leveraging automation and AI.

Bona Bai, PhD

Bona Bai, PhD

Robotics Lead

15 years building autonomous systems across Amazon (Astro), Google Wing, and Figure AI. Full-stack autonomy from sensor fusion and perception through to real-time decision making.

Prof. Tomoya Nakamura

Prof. Tomoya Nakamura

Research Advisor

Professor at The University of Osaka and Visiting Professor at Stanford University. Computational Imaging, Computer Vision, AI Optics.

Join us

We're hiring. Work on problems that matter.

Founding Business Member

Build Yodo Labs from the ground up. Get onto customer sites across infrastructure, construction, and heavy industry, and turn early conversations into real deployments alongside our engineering team.

Expressions of Interest

Don't see your role here? We're always looking for strong people. Tell us what you'd bring and where you'd want to contribute.