Edge AI Spin-out Project from Germany

Bringing Edge AI Prediction
to Industrial Applications.

Lightweight, high-accuracy real-time prediction AI that runs on small devices.
No cloud required. No GPU required. Predict unmeasurable values,
detect anomalies, and forecast remaining useful life from existing sensor data.

Spin-out from
DLR German Aerospace Center

Entrox Systems is a spin-out project from the DLR German Aerospace Center. We are developing reservoir computing technology from the research lab for real-world industrial applications.

Based in Germany, developing cutting-edge prediction technology for industrial sites worldwide.

Predict hard-to-measure values
in real time, on the factory floor

Industrial sites generate diverse time-series data from sensors and equipment. Our AI learns patterns from even small amounts of time-series data, enabling real-time prediction of hard-to-measure values, anomaly detection, and remaining life forecasting.

Edge AI

Runs on small, low-cost devices. No cloud connection required — your data stays on-site.

Instant Training

Training completes in seconds. Quickly adapts to new environments and starts running immediately.

Lightweight Precision

Delivers performance comparable to large-scale AI on low-cost hardware.

Custom Optimization

Builds optimized models per site and use case. Flexibly handles high-mix, low-volume production.

Test our AI on your data — results in as little as 1 day. No GPU or cloud setup required.

Three Prediction Solutions

Addressing the full spectrum of industrial prediction challenges.

SOLUTION 01

Soft Sensors

Predict what you can't measure

Real-time prediction of hard-to-measure parameters from existing sensor data

Benchmark Results
  • HVAC refrigerant pressure prediction (R²=0.92)
  • Pharmaceutical tablet hardness prediction
  • Semiconductor etching process pressure prediction
  • Electricity demand forecasting & heat pump COP estimation
Use Cases
  • Chemical reactor temperature & concentration prediction
  • Digital twin state estimation foundation
  • Food & paper process quality prediction
SOLUTION 02

Anomaly Detection & Predictive Maintenance

Spot what's different before it fails

Learns "normal behavior" of equipment and detects early signs of anomalies

Benchmark Results
  • HVAC refrigerant leak detection
  • Compressor degradation detection
  • Heat exchanger fouling detection
  • Pump vibration anomaly detection (AUC=0.85+)
Use Cases
  • Rotating equipment & plant monitoring
  • Real-time inference on IoT devices
  • Preventive maintenance for energy equipment
Only normal-state data required for training
SOLUTION 03

Remaining Useful Life (RUL)

Predict when it will fail

Predicts remaining useful life from sensor data to enable planned maintenance

Benchmark Results
  • Battery SOH prediction (R²=0.92)
  • Truck component RUL prediction
  • Turbine engine remaining life prediction
Use Cases
  • Bearing & gearbox lifetime prediction
  • Battery & energy storage degradation prediction
  • Wind turbine structural fatigue prediction
Prevent unplanned downtime, reduce maintenance costs

Value for Your Business

Advanced Quality Control

Real-time prediction of hard-to-measure quality parameters. Reduce analysis wait times and enable early defect detection.

Predictive Maintenance

Early detection of anomalies in pumps, compressors, and more. Prevent unplanned downtime and reduce maintenance costs.

Process Optimization

Predict quality and efficiency from operating conditions. Reduce trial-and-error and digitize expert knowledge.

High-Mix Adaptability

Build individual models per product and line in minutes. Flexibly adapt to high-mix, low-volume production.

The Edge AI Advantage

No expensive servers or cloud subscriptions needed

Data stays on-site — enhanced security and privacy

Fast deployment — models applied per site in minutes

Scalable — easy rollout across multiple sites and lines

Our Team

Tamon Nakano

Tamon Nakano, Ph.D.

Co-founder

AI, Mechanical Engineering, Business Development

Daniel Köglmayr

Daniel Köglmayr

Co-founder

AI, Algorithm Development

Sebastian Baur

Sebastian Baur

Co-founder

AI, Software Development

PoC Partners

We co-develop solutions through proof-of-concept projects with leading technology partners.

Research Track Record

Our founding research group is based at DLR, the German Aerospace Center, conducting international collaborative research.

6
Patents
20+
Peer-Reviewed Papers
350+
Academic Citations

Published in IEEE, Elsevier, Springer, and other leading journals.
An integrated development pipeline from research to real-world deployment.

Start with a Small-Scale Pilot

We conduct small-scale pilot projects using your real data to demonstrate effectiveness in your actual environment. Get in touch to discuss your use case.

Test our AI on your data — results in as little as 1 day. Our AI is extremely lightweight — no GPU or cloud infrastructure needed.
Impressum
Published by: Entrox Systems, Kehl, Germany.
Represented by: Tamon Nakano, Kinzigstraße 59, 77694 Kehl, Germany. Email: info@entrox-systems.com
DLR Spin-out project