Managed by us
Fastest to stand up. Streams are processed in your region, models are managed for you, and nothing is shared across tenants.
VRTLST Edge v2 ships with on-device inference at 38 ms
Read moreThe platform
Six products that were designed together rather than acquired separately. Vision is the engine, Edge is where it runs, Inspect, Guard and Watch are what it does for you, and Command is where you steer the whole thing. Nothing to stitch together.
Architecture
Read it from the bottom. Everything rests on the edge runtime, and everything above shares the same model.
Control plane · fleet · access · audit
Defect detection
Safety & security
Monitoring
Detection · segmentation · tracking · classification
On-device runtime · cameras · gateways · air-gapped sites
Layer by layer
The reason a defect model and a safety model agree about what they are looking at is that they are the same model, looking once.
Book a demoWhere it runs
The same model artefact runs in all three. Moving between them is a deployment decision, not a rebuild.
Fastest to stand up. Streams are processed in your region, models are managed for you, and nothing is shared across tenants.
A plant server runs the stack behind your firewall. Command reaches it over a link you control, and frames never leave the site.
Inference runs on the camera. The air-gapped configuration has no outbound connection at all — models arrive by hand and stay put.
How it goes
Every deployment we have run follows the same shape. Nothing here needs a data-science team on your side.
Send us recorded video from a single line. We come back with what the model found in it — before anyone signs anything.
We stand up a pilot on the line that hurts most, using the cameras already mounted. Operators start correcting it on day one.
The verdict moves from a dashboard to the PLC, the MES or the alerting channel your shift leads actually watch.
The model and its thresholds travel. Command stages the rollout so a bad version never reaches more than one line.
Corrections accumulate, drift gets flagged, and new defect classes are a fine-tune rather than a project.
Security
Vision projects die in security review. These are the four answers that get asked for every time, and they are architectural rather than contractual.
Not on any plan, not in any deployment. There is no shared model that has seen your floor.
Not a workaround. Models arrive on media, inference runs locally, and nothing dials out.
Blurring happens before the frame is written anywhere. There is no identity index to leak.
SAML and OIDC, roles down to the site, and an exportable record of every change.
Integrations
VRTLST is a sensor and a decision, not a replacement for your control system. It writes into what you have.
Control layer
Plant systems
Data & messaging
People
The products
Detection, segmentation, tracking and classification from a single deployment — no per-camera retraining.
Explore VisionSub-millimetre surface inspection at line speed, with a decision the PLC can act on.
Explore InspectDraw a boundary once. Guard enforces it on every shift, on every camera, without a human staring at a wall of screens.
Explore GuardContinuous monitoring across every line and every site, with alerts that reach the right person in seconds.
Explore WatchA sealed edge-AI device that runs the full VRTLST stack on-board. Single PoE cable in. Decisions out.
Explore EdgeThe control plane for your whole visual estate — deployments, model versions, alerts, access and audit.
Explore CommandGet started
Send us one hour of footage from one line. We'll come back with what the model found in it.