NEXEGRA
- What we did
- Industrial AI platform that connects OT and IT systems across production lines, delivering real-time monitoring, AI-driven optimisation and predictive maintenance. Deployed at Van Dongen Bakkerijen across 3 production lines with 146 sensors.
- Type of work
- OT/IT Integration & Industrial AI
- Industry
- Food production & manufacturing
- Done by
- Nexegra & Atlantis Digital
Nexegra is an industrial AI platform that bridges the gap between operational technology (OT) on the production floor and enterprise IT systems. Deployed at Van Dongen Bakkerijen — a mid-size industrial bakery running 3 production lines with equipment from Siemens, Beckhoff and Allen-Bradley.
What used to be a black box of isolated machine data is now a real-time intelligence layer. 146 sensors feed 2.4 million data points per day through Nexegra's integration platform — delivering live OEE dashboards, AI-driven process optimisation and predictive maintenance alerts that prevent downtime before it happens. Result: €127,400 annual savings and OEE up from 71% to 89% in 14 weeks.
Production
Nexegra's
OEE
Legacy
REAL-TIME
From
146
Sensors
2.4M
Data
23ms
Edge-to-dashboard
99.97%
Platform
PROTOCOL-AGNOSTIC
Van Dongen's production floor runs on three different PLC platforms — Siemens S7-1500, Beckhoff CX and Allen-Bradley. Nexegra's integration layer speaks all protocols: OPC-UA, MQTT and REST API. Edge compute nodes handle protocol translation and data normalisation on-site. The result: one unified data stream from machines that were never designed to talk to each other.
AI
Generic monitoring spots out-of-range values. Nexegra's AI understands that a 2°C oven deviation during proving affects dough hydration 20 minutes later. Domain-specific ML models correlate parameters across the entire line — from dough viscosity in the kneader to packaging seal integrity — identifying optimisation opportunities that human operators miss.
MEASURABLE
Hard
89%
OEE
67%
Reduction
€127K
Annual
PREDICTIVE
Nexegra's ML models predict equipment failure up to 72 hours in advance with 91% confidence. At Van Dongen, the system flagged a cooling spiral compressor degradation that would have caused 8 hours of unplanned downtime — saving €2,400 in a single intervention. Maintenance shifted from reactive to planned: scheduled during night shifts, zero production impact.