Interactive
prototype
- Three.js scene and synthetic signals
- Layout edits and sampled checks
- Cycle and cost assumptions
NOUS
Digital-twin engineering for manufacturing decisions.
On site. One critical cell first.

The problem · Unplanned downtime
Reported cost of an idle line.
About 65% longer than five years earlier.
In just-in-time production, a stopped cell can delay the next plant.
Siemens, The True Cost of Downtime 2024. Large-plant benchmarks; costs vary by plant.
Customer & market · Ontario first
Tier 1 / Tier 2 JIT suppliers · 100+ employees · Robotic lines · Retooling ahead
1,200 businesses × CA$85k
350 plants × CA$85k
15–25 plants × CA$85k
Illustrative annual revenue scenario, using assumed CA$85k blended annual spend per account. Counts require qualification and deduplication.
The service · Proposed forward-deployed model
CA$35k · 2–4 weeks.
Check clashes and downtime exposure before retooling.
Partner on scans and controls.
Reuse asset models and CAD-to-OpenUSD workflows.
CA$85k–150k+ deployments.
Add lines and ongoing support once value is established.
Specialist partners for scanning and controls.
Proposed delivery and pricing; scope agreed first. OpenUSD ingestion and NVIDIA integration are next-stage engineering work.
01 / Plan · Industry case · BMW
FactoryExplorer brings building, equipment and logistics data into one planning environment.

OpenUSD + custom Omniverse Kit tools for measurement and collision review. The saving is a projection, not our customer result.
02 / Retool · Industry case · Schaeffler
Simulate robotic tasks and production configurations before deploying changes.
OpenUSD, simulation and virtual commissioning support the application. This case concerns robot/process reconfiguration; no whole-plant retooling saving is assumed.
03 / Improve · Industry case · Wistron
Production data and process simulation reveal bottlenecks and guide layout changes.
Company-reported, March 2024. The metric is specific to the bottleneck line; it is not a plant-wide OEE result or a forecast for our service.
Working browser prototype

Follow a missing scanner input through diagnosis and repair.
Synthetic scene and signals. This is a browser demonstration, not a connected Omniverse or PLC deployment. Drag the sensor in Realign to restore detection.
Our working prototype / Retooling


Drag equipment · adjust bolting and speed · compare options
Our synthetic scene and custom kinematics. Clearance is sampled; timing and installation costs use explicit demo assumptions.
Proposed production architecture
CAD / scan geometry
Robot and tool specifications
OpenUSD assets
Layers, references, alternatives
Omniverse libraries / Kit
Robot and process models
Browser comparisons
Engineering validation
Later: operational dataTimestamped machine events, counts and quality records update the baseline.
Asset preparation · process logic · integration · validation · cost assumptions
NVIDIA supplies modular building blocks. Geometry needs behavior and calibrated inputs; a GPU-backed runtime would be separate from this static website.
Technology readiness / Our evidence
NVIDIA runtime integration, live PLC data and measured customer outcomes remain future work. Current nominal capacity is not measured OEE.
Commercial model · Pricing hypothesis
One scoped change · 2–4 weeks.
Clash review and downtime exposure.
Scoped full-line deployment.
Ongoing retainers quoted separately.
Grant and loan eligibility is project-specific. No subsidy is assumed in the customer business case.
Proposed prices, not booked revenue or validated willingness to pay. Final scope, hardware, licensing, partner costs and acceptance criteria agreed before work starts.
The ask
Introduce an owner or plant engineer.
Start with feedback, then one bounded audit.
After discovery, bring together simulation,
robotics and controls expertise to deliver it.
Seeking introductions and engineering collaborators. No customer engagement or partnership is claimed.