Software
AI Vision Tracking
Classification, object detection and multi-camera tracking for production and logistics - without barcodes or RFID. From site mapping and simulation to trained weights and live tracking on the floor.
See it in action
Overhead cameras track pallets with a handover from one camera to another
The bounding box follows the pallet as it moves along the conveyor, crossing entry and exit zones that trigger events for your PLC, MES or WMS.
Overview
What 3ops AI Vision does
A flexible AI detection and tracking platform built around a best-in-class tracking engine, extended for multi-camera tracking of the same object across MES and WMS environments.
Detection & classification
Identify objects, pallets, cartons, people or assets in the frame in real time.
Multi-camera tracking
Track the same item across multiple camera streams for redundancy and full path coverage.
Zones & rules
Draw polygons for entry/exit, buffers and quality zones. Trigger events when objects cross them.
Anomaly detection
Catch stacking, wrapping or labelling anomalies before they reach the customer.
PLC & MES/WMS integration
Signal PLC queues, MES and WMS so the process can halt, divert or confirm automatically.
Edge deployment
Runs at the edge on NVIDIA Jetson Orin - no cloud round-trip, low latency, resilient to network drops.
Project stages
From floorplan to live tracking
A clear delivery path - flexible enough to fit a one-day training or a full turn-key project.
- Stage 01
Map the area
We start from your floorplan and on-site measurements to map the area where tracking is needed - lines, conveyors, racks, zones, lighting and obstructions.
- Stage 02
Simulate camera placement
Using our 3D simulation we evaluate field-of-view, occlusion, mounting points and overlap, then agree on the exact camera positions with you.
- Stage 03
Hardware advice
We advise on the right hardware mix: NVIDIA Jetson Orin compute, fisheye / IP cameras, PoE switching and power - sized to your scope and budget.
- Stage 04
Assembled cabinet (optional)
We deliver a pre-assembled Rittal (Inox) cabinet with the AI compute, PoE, wiring scheme and documentation. You only place the cabinet and cameras.
- Stage 05
Training day or full project
Choose your delivery model: a one-day training so your team can do the rest themselves, or a full turn-key delivery by 3ops.
- Stage 06
Upload training data
Upload images from your site to our portal. 3ops handles the annotation while you keep full ownership of data and weights.
- Stage 07
Decide & train
You decide which classes to train and launch training runs from the portal - getting back versioned weight files when ready.
- Stage 08
Self-service configuration
Configure cameras, zones and tracking rules yourself in the 3ops Vision Manager using the weight file generated in our portal.
Option A
One-day training
We deliver the hardware and cabinet, train your team in a single day on the 3ops Vision Manager, annotation portal and configuration workflow - your team takes it from there.
Option B
Full turn-key project
3ops handles everything: mapping, simulation, hardware, cabinet, annotation, training runs, configuration and go-live - you receive a working tracking system.
Your portal, your data
Annotation, training and configuration - self-service
You upload training data to the 3ops portal. 3ops annotates. You pick which classes to train, kick off training runs and download the resulting weight file. Then you configure cameras, zones and rules yourself in the Vision Manager. You own the data and the trained weights.
Ready to transform your operations?
Talk to a 3ops expert about your manufacturing or logistics challenge.