Xenith Spatial
Cameras to living twin to human-approved action.
Xenith Spatial unifies computer vision, a spatial data model, and HITL optimization on existing plant and warehouse cameras. Land with vision-lite incident reduction; expand into predictive twin and optimizer tiers when pilots prove signal quality. LLMs do not route forklifts; rules and approvals do.
Deployed as a scoped engagement for your stack - not a self-serve trial or live sandbox.
Why teams look for this
Problems this pattern addresses
Alert dashboards without action
CCTV and CV vendors stop at notifications. Ops still chase clips in silos.
Static twins
3D models without live detections cannot support routing or safety decisions.
LLM creep into physical routing
Models must not route forklifts. Rules and approvals own motion.
Where it sits in your stack
How it works
Vision Lite
PPE, near-miss, intake audit. Land here first (4–8 weeks).
Spatial Twin
Living canvas with freshness SLOs and predictive alerts.
Optimizer
Throughput proposals only after precision earns Assist.
Governed AI
Safety, audit, and rollout rules
Read the Governed AI doctrine for how Xenqube separates model output from money and safety actions.
Do not sell Tier 3 first
Land Vision Lite. Expand when measured precision and operator trust justify optimizer proposals.
HITL on every automated action
No silent closed loop on people or equipment movement.
Proof and measurement
Workshop and shadow-trial metrics are labeled when synthetic. Live SLAs require a named pilot agreement.
| Metric | Target / demo | Notes |
|---|---|---|
| Pilot length (Vision Lite) | 4–8 weeks | Typical SOW when cameras exist |
| Phase-3 targets | Aspirational | Precision and downtime bands are labelled targets, not measured claims |
| LLM forklift routing | Forbidden | Rules + approvals only |
Who this is for
Use cases
Rollout
Phased trust model
Coverage, lighting, privacy, and event spine.
Alerts with human disposition.
Spatial model freshness and predictive signals.
Approved proposals only after precision gates.
What's included
Typical engagement deliverables
Scoped to your stack and compliance needs. We deploy the working system into your environment and hand over runbooks - not a sandbox login.
Synthetic or pilot camera feed into alert rules
Twin UI with person/vehicle classes
HITL approval on any automated action
Tier expansion plan tied to measured precision
Questions buyers ask
Frequently asked questions
- Do we need new cameras?
- Often no. We start with an audit of existing CCTV. New cameras only when coverage or quality fails the pilot.
- Will AI move equipment autonomously?
- Not without Assist-phase approval design. Tier 3 proposals stay behind human gates.
Related products
More in Industrial AI