01
Coating & foil defects
Pinholes, scratches, foreign particles, edge burrs, and coating defects on aluminium foil, copper foil, lithium battery electrode sheets, and coating lines.
Industrial vision · Research preview
MotionEye
Event-based industrial vision module.
MotionEye is an event-based industrial vision module that detects tiny defects and abnormal events on high-speed production lines — and works with Zoplis DataBox to turn every critical event into traceable, audit-ready evidence.
Designed for aluminium foil, copper foil, lithium battery coating, precision manufacturing, and high-speed counting applications.
Research preview · GA Q1 2027
Application areas
Where MotionEye fits.
02
High-speed counting
High-speed counting of bottles, batteries, tablets, and packaging parts — including missed-count and double-count prevention.
03
Event traceability
Capturing abnormal events on high-speed production lines, supporting incident traceability and quality evidence retention.
04
Future expansion
Pharmaceutical packaging, food packaging, semiconductor materials, and precision manufacturing.
How MotionEye works
See, classify, lock, trace.
- 01
See
Event camera captures what frame cameras miss. Microsecond-level resolution. >120 dB dynamic range. No motion blur, no dropped frames.
- 02
Classify
Edge AI applies three-state decision in real time — KEEP / REJECT / DEFER. No forced binary classification.
- 03
Lock
DataBox locks each event into the evidence chain — timestamp, machine state, and process parameters at that exact moment.
- 04
Trace
Any defect, any time later, retrievable. 14 months after shipment? Pull the exact second, exact meter, exact machine state, exact operator. In 30 seconds.
How it works with DataBox
MotionEye sees + DataBox records
MotionEye captures the event — pinholes, scratches, coating defects, counting errors. DataBox provides the audit trail — timestamps, machine status, sensor data, alarm records, hash signatures, trusted timestamps, audit reports. The customer knows not only that something went wrong, but when, where, under what conditions, and how to verify it later.
Technology directions
Five engineering tracks under v1.1.0.
MotionEye is not a packaged off-the-shelf camera. It is a vision module engineered for the specific physics, optics, and data demands of high-speed industrial inspection.
Physics AI
Combining material constitutive equations with AI to explain defects in aluminium foil, copper foil, cold rolling, and coating processes.
Online Learning
Detecting production-line drift to prevent models from becoming inaccurate over long-term use.
Equivariant CNN
Few-shot defect recognition that reduces the need for large amounts of labelled customer data.
Synthetic Defects
Procedurally generating defect samples — pinholes, scratches, inclusions, wrinkles, edge burrs, streaks, dust, and glare.
Spiking / Event Processing
Preparing for future integration between event cameras and neuromorphic hardware.
Release status
Now in research release. Full launch Q1 2027.
MotionEye v1.1.0 is currently a research release. The next stage will be pilot tests on real production lines, validating detection rate, false-alarm rate, system stability, and reporting capability. Full commercial release is planned for Q1 2027.
Talk to us about MotionEye.
For supplier conversations, pilot programs, or early-access partnerships, get in touch — we respond personally.