NexTex AI
Industrial machine intelligence
The machine wakes.
Vision comes online.
Every meter, understood.
The problem
Most defects are found at rolling or by the customer — hundreds of meters after they were produced. The meters in between are already cost, waste and CO₂.
PLC signals, recipes, batch context and operator knowledge live in separate systems. No one can connect a machine event to its quality and sustainability outcome.
CSRD, the EU Green Deal and Digital Product Passports ask for machine-level evidence. Spreadsheets assembled once a year cannot carry that burden.
Manufacturing does not need another dashboard. It needs machines that can see, explain and prove what they do.
The production lifecycle
01 · Knitting
Loop formation is watched at needle resolution — defects are caught as they are knitted, not at rolling.
How NexTex AI works
Industrial cameras, retrofit sensors and PLC / SCADA taps capture every machine event — without stopping the line.
Edge AI models detect defects and contextualize machine behavior in real time, meters from where it happens.
The platform proposes process improvements. Operators review, approve or decline — no autonomous machine control.
Every event lands in the sustainability intelligence database and becomes a traceable, audit-ready KPI.
AI defect intelligence
Operator alert · line 2
Recurring needle line, machine M-12, needle 214. Root cause flagged — inspection recommended before the next batch.
Each finding carries a class, a confidence score, machine context and position on the web — so quality teams act on causes, not symptoms. Validated findings feed the continuous learning loop.
AI process intelligence
NexTex AI correlates machine behavior with quality and resource outcomes, then recommends the parameter window that produces first-quality fabric with less energy, water and chemistry.
Human-in-the-loop by design
NexTex AI recommends. Operators decide. No process parameter changes without explicit, logged operator approval — ever.
Recommendation · #241
Stenter M-07 — reduce drying zone 3 by 4 °C. Estimated −6% gas per batch; ΔE stability unchanged within tolerance.
Industrial data intelligence
Machines
Context
Environment
The sustainability intelligence database
Every machine event is connected to its quality, process and sustainability outcome — one queryable record of how fabric actually gets made.
Sustainability intelligence & reporting
Reporting is not a dashboard bolted on at the end — it is the product of every sensed event. Numbers below are illustrative pilot telemetry; in production, each figure drills down to the machine events that produced it.
−8% vs target
−12% YoY
scope 1+2 basis
−0.6 pt since pilot
graded per meter
root-caused
event-based OEE
per-batch drift
KPIs structured along CSRD reporting logic, generated from machine events rather than estimates.
Batch-level material, process and impact data organized so DPP requirements can be met as they land.
Energy, water, chemical and waste intensity tracked continuously against reduction targets.
Traceable measurement chains and methodology notes prepared for third-party review.
Every KPI traces to the machine events behind it — drill down from report line to timestamp.
Management, operations, quality and sustainability reports exported as structured PDF packs.
Every KPI traces to a machine event. Every report can be defended in an audit.
Products
Product 01
AI-powered machine monitoring for textile manufacturers starting digital transformation — see your quality and resource footprint in weeks, not years.
Start monitoring · retrofit install
Product 02
The complete industrial intelligence platform — process optimization, digital twin and audit-ready sustainability intelligence across factories.
Scale intelligence · factory-wide
Fabric upgrades to Fabric Pro without replacing hardware — the same cameras, sensors and edge devices carry both.
Industrial AI architecture
Inference runs at the edge, meters from the fabric. The cloud holds the intelligence and the evidence. Every layer is replaceable without stopping production.
01
Machine layer
Industrial cameras
Retrofit sensors
PLC / SCADA
Machine events
02
Edge layer
Edge AI inference
Event capture
On-prem buffering
Line-speed latency
03
Intelligence layer
Defect models
Process models
Digital twin
Sustainability database
04
Application layer
Dashboards
Operator alerts
Audit-ready reports
API
Data sovereignty by default: fabric imagery can stay on premises — only events, KPIs and model updates cross the boundary.
Continuous learning loop
Models flag defects and process drift in real time.
Operators confirm or correct every finding — human-in-the-loop.
Validated findings become training signal for the next model.
Improved models roll to the edge with full version history.
Detection quality and false-alarm rates are tracked like any KPI.
The loop is closed by people, not around them: operator validation is what turns detections into ground truth.
Pilot environments
A NexTex AI pilot is an engineering engagement with defined success criteria — measured against your fabric, your machines and your baseline, before any rollout decision.
Roadmap
Now
Next
Later
Team
NexTex AI is a European DeepTech team. We publish claims we can defend — the same standard we apply to every report the platform produces.
Engineers who ship models that hold up at line speed, in mill light, on real fabric.
People who have run knitting, dyeing and finishing lines — and know where quality is actually lost.
Builders of measurement chains and reporting tooling designed to survive an audit.
We hire engineers who have stood next to the machines they model. careers@nextex-ai.com
Contact
Tell us about your lines — machines, fabrics, current inspection setup — and we will scope a pilot with measurable success criteria.
