AIthon + I-Edge01
Training and edge deployment
A manufacturing appearance-inspection platform covering projects, data, training, evaluation, model management, inference and equipment interfaces.
Status: Product material exists; metrics need reports
INDUSTRIAL INSPECTION
From seeing defects to understanding process.
GENSTAR combines optics, mechanics, algorithms, software and process knowledge to sense appearance, dimensional, assembly and process anomalies.
FIELD STRUCTURE
Platforms, modular equipment and project solutions
The November 2025 material provides product names, equipment forms and functional scope. Commercial status, production status and performance figures still require owner and test evidence.
SEMIVUE02
Wire-bond 2D/3D inspection
Inspect frames, ICs, devices, wires and bonds, classify defects and analyse results for in-line or off-line workflows.
Status: Equipment form confirmed; specifications need alignment
INSIGHTX03
Six-side component inspection
Combine multi-side imaging, AOI and AI for automatic electronic-component appearance inspection and result analysis.
Status: Equipment material exists; commercial status pending
AIBOX04
Edge AI upgrade for existing AOI
Connect existing AOI and PLC systems with minimal intrusion to filter false calls, add defect recognition and classification.
Status: Formal compatibility list pending
FLEXIBLE CELL05
Vision soldering and inspection cells
Combine robotics, optics, positioning correction, soldering control, six-side inspection, traceability and business feedback.
Status: Prototype imagery exists; production status pending
PROJECT SOLUTIONS06
Project-based vision solutions
Cover OCR and print quality, assembly sequence, 3D profiling, dimensional measurement, safety and factory-wide monitoring.
Status: Assessed and configured by scenario
ANSWER ENGINE
Common questions and boundaries
Direct answers to early-stage questions, including boundaries that cannot be assumed.
01Why does industrial vision start with defect definition?+
Defect definitions determine samples, optics, algorithms and acceptance. Without a shared boundary, even a high model score cannot create a stable production decision.
02Can AI fully replace rules and human review?+
Usually not. Robust systems combine rules, learned models, anomaly detection and human review according to operational risk.
03Is photon-counting a mature GENSTAR product?+
No. It remains an advanced-imaging exploration and will only be presented as a product after status, evidence and delivery boundaries are confirmed.
04Can Semivue, InSightX or AIthon figures be treated as project guarantees?+
No. Current material supports the product names and capability framework, but precision, takt, detection and false-call figures depend on parts, optics, defect standards and formal test evidence.
NEXT STEP
Discuss your challengeBegin by making the field challenge clear.
Share the current process, constraints and intended outcome. We will assess the smallest verifiable starting scope.