NexTex AI
Industrial machine intelligence
The machine wakes.
Vision comes online.
Every meter, understood.
Design targets — validated per mill during pilots
The problem
Across peer-reviewed studies, manual fabric inspection catches only 60–75% of defects — and accuracy falls as the shift wears on.1 Every missed meter travels downstream as cost, waste and customer claims.
From around 2028, textiles sold in the EU are expected to carry a Digital Product Passport built from production data.4 Brands are already asking suppliers for batch-level evidence — long before regulators ask them.
Manufacturing does not need another dashboard. It needs machines that can see, explain and prove what they do.
Industrial pilot partners
Terrot GmbH
Circular knitting machine manufacturer — Chemnitz, Germany
Nuryıldız Tekstil
60+ machines, circular knitting & finishing — Çorlu, Türkiye
Ribana Tekstil
60+ circular knitting machines — Çorlu, Türkiye
ONE WARE GmbH
Hardware-optimized AI deployment & FPGA edge inference — technology partner, Germany
Pilot preparation — not customers, not production deployments
The production lifecycle
01 · Knitting
Designed to watch loop formation at needle resolution — catching defects 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 KPI — audit-ready by design.
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 · illustrative example
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 · in development with pilot partners
Designed so every machine event connects to its quality, process and sustainability outcome — one queryable record of how fabric actually gets made.
Sustainability intelligence & reporting
Five numbers tell the story. Everything beneath them stays traceable to machine events — each figure designed to drill down to the events that produced it.
Illustrative design targets — validated against each mill's own baselines during pilots
Energy
0.42
kWh/kg
−8% vs target
Water
62
L/kg
−12% YoY
CO₂e
1.9
kg/kg
scope 1+2
Waste
1.8
%
−0.6 pt since pilot
Quality
98.5
%
first quality
Machine efficiency
84%
event-based OEE
Rework
0.9%
root-caused
ΔE color stability
0.4 avg
per-batch drift
Sustainability report
Illustrative
Designed so every figure traces to machine events · Export PDF
One page, five numbers, a trend — and behind every line, the machine events that produced it. Drill from report figure to timestamp when the question comes. NexTex AI prepares the evidence; certification remains with accredited bodies.
Products
Product 01
AI-powered machine monitoring for textile manufacturers starting digital transformation — see your quality and resource footprint in weeks, not years.
Pilot scope
Retrofit install · pilot pricing on request
Product 02
The complete industrial intelligence platform — process optimization, digital twin and sustainability intelligence across factories.
Pilot scope
Planned
Sequenced with pilot partners
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 will 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.
Current status
Letters of Intent signed with three industrial partners. First pilot mills onboarding for 2026–27. No commercial deployments yet — and we say so.
Roadmap
Now
Next
Later
Founders
We publish claims we can defend — the same standard we apply to every report the platform produces.
Co-Founder — Backend Infrastructure & Full-Stack Engineering
Co-Founder — System, Edge & Data Architecture
We’re hiring a Founding Software Engineer. 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.
