VEXPLOR
KO/ ENRequest a Diagnosis
CUSTOMER CASES

Sites we can nameand stand behind.

From two Korea MSS Autonomous Factory Program sites to MES and PLM builds, we show you exactly what we built and with what technology. Planned figures, we label as planned.

20+ PROJECTS · 2 NATIONAL AUTONOMOUS-FACTORY PROGRAMS · 6 MFG-AI PROGRAMS · 8 GOV PROGRAM TYPES
01PROJECT LEDGERNAMED · DATED · STACKED
NO.Customer · IndustryScope · Technical stackPeriod
C-01SpeefoxCapacitor cases · press mass productionKorea MSS Autonomous Factory Program, selected 2025 — full-stack perception-decision-execution closed loopSTACKMES · UNITY 3D WEBGL TWIN (100MS SYNC) · AI AUTONOMOUS CONTROL · AAS (BASYX) · EDGE
SCALE186 PRESSES · 25 AMR · 6 EMS SIMULTANEOUS SYNC [VERIFIED] · DES ACC 85% ACCREDITED TEST [VERIFIED]2025.06–2027.06
C-02ChexcarUsed-car remanufacturingKorea MSS Autonomous Factory Program, selected 2026 — transplanting the Speefox stack (proving reproducibility from high-volume to high-mix)STACKMES + SCADA · DIGITAL TWIN/CPS · IOT — SAME STACK AS SPEEFOX2026.06–2028.05
C-03Jeil GlassInsulated tempered glassMES + Order-based PLM — a single data flow from order to design to production to shipmentSTACKMES · ORDER-BASED PLM · REAL-TIME EQUIPMENT MONITORING
RESULT100% AUTOMATED LOT TRACEABILITY [CUSTOMER RESULT]2024–2025
C-04NKIDAluminum die castingPhase 1 MES → Phase 2 Manufacturing-AI Program — extended to AI on the same standard tablesSTACKMES · PREDICTIVE MAINTENANCE (LSTM AE+CNN) · QUALITY PREDICTION (LSTM-CNN+ATTENTION) · PROCESS OPTIMIZATION · SAAS MES/FEMS DEMO · X-RAY 3D CT INTEGRATION2023.09–2026.06
C-05IntopsElectronic componentsPLM (ProSearch) · unified search · sales management — enterprise-wide search across design assetsSTACKPLM · UNIFIED SEARCH · SALES MANAGEMENT MODULES2024.11–2025.07
C-06TopsealComponent manufacturingWin-Win Program (Samsung Electronics) MES enhancement — stronger data collection and quality controlSTACKMES ENHANCEMENT · EQUIPMENT DATA COLLECTION · QUALITY CONTROL2025.07–2026.02
C-07JustemSemiconductor components (N2 PURGE)Manufacturing-AI Program — multi-plant procurement/materials AI agents, integrated via ontology without replacing the existing ERPSTACK5 PROCUREMENT/MATERIALS AI AGENTS · K-SYSTEM ERP + PDM ONTOLOGY MAPPING2026.07–2027.03
C-08Sunyoung KoreaSMT electronics componentsManufacturing-AI Program — manufacturing-process agents and vision inspection, tied to Hyundai SQ certification readinessSTACKMANUFACTURING-PROCESS AI AGENT · AOI VISION INSPECTION2026–2027
C-09RPSAir-bearing spindlesOrder-based PLM and equipment-build management → full enterprise system build + Manufacturing-AI Program — replacing Douzone iCube ERP with VEXPLOR (keeping only HR and accounting on Douzone Amaranth)STACKERP REPLACEMENT (ICUBE → VEXPLOR) · ORDER-BASED PLM · EQUIPMENT-BUILD MANAGEMENT · 5 PROCUREMENT/MATERIALS AI AGENTS
KEEPHR & ACCOUNTING — KEPT ON DOUZONE AMARANTH; EVERYTHING ELSE BUILT BY WACE2025.08–2027
C-10AlpetPET film laminatingAI laminating quality prediction — from process data collection to prediction and optimizationSTACKQUALITY PREDICTION AI · MES · IOT GATEWAY · PROCESS OPTIMIZATION AI2025.10–2026.06
C-11Speefox Geomdan PlantPlating (heat-conduction plates)Manufacturing-AI Program — plating-line agents, an expansion order following the autonomous-factory demonstration at headquartersSTACK5 PLATING-LINE AI AGENTS2026–2027
C-12HurumHealth functional foodInter-ministry Program smart factory — production management on a food-GMP quality systemSTACKMES · FOOD-GMP COMPLIANCE · BUILT ON AAS STANDARD CAPABILITY2026–2027
C-13Daejin IndustrialMES build (Standard Government Program) — production management, integrated management, equipment interfacesSTACKMES · INTEGRATED MANAGEMENT · EQUIPMENT INTERFACES2025.05–2026.09
C-14Ilshin AutoclavePLM build — running in parallel with a Win-Win Program (Samsung Electronics, foundational)STACKPLM · WIN-WIN SMART-FACTORY FOUNDATION2025.06–2026.12
C-15Company S K-System ERP ecosystem — via official partnerK-System ERP add-on POP build and supplySTACKPOP (POINT-OF-PRODUCTION) · K-SYSTEM ERP INTEGRATION2026.06–2027.01
-15%Company H (die casting) — defect-rate reductionCUSTOMER RESULT
99%Company K (automotive electronics) — inventory accuracyCUSTOMER RESULT
-20% · -90% · -15%Defect rate · data-processing time · process operating costEXPECTED IMPACT — NOT A CUSTOMER RESULT
Planned program target figures are not shown · VERIFIED = field demonstration complete · CUSTOMER RESULT = measured by the customer
CASE C-01 · KOREA MSS AUTONOMOUS FACTORY PROGRAM — SELECTED 2025

Speefox —
full-stack closed-loop demonstration

Press industry mood imageINDUSTRY MOOD IMAGE
ProblemA capacitor-case mass-production plant running 186 presses. Plenty of equipment, but the data was scattered, and there was no system to see and operate the whole line as one.
Technical stack (2025.06–2027.06) PERCEPTION — equipment and vision data collection, AAS (BaSyx) asset standardization, EDGE gateway
DECISION — DES-simulation-based AI decisioning, stepwise AAF control handover (starting from Shadow Mode)
EXECUTION — AI → PLC direct control, command validation at the EDGE node
TWIN — Unity 3D WebGL digital twin, 100ms floor sync in rollout
Demonstrated results — to date [VERIFIED]Simultaneous synchronization of 186 presses, 25 AMRs and 6 EMS units demonstrated (the AMRs and EMS are the customer's existing equipment, integrated under AI supervision) · DES simulation accuracy 85%, certified by an accredited testing body · AI → PLC direct control demonstrated — on physical equipment.
NextThe project runs through June 2027. The same stack has been transplanted to Chexcar (a used-car remanufacturing site selected in 2026), proving reproducibility across different process types — high-volume and high-mix.
CASE C-03 · MES + ORDER-BASED PLM

Jeil Glass —
one flow from order to shipment

Glass industry mood imageINDUSTRY MOOD IMAGE
ProblemInsulated tempered glass manufacturing — every order has different specs, and LOT tracing relied on paper, so tracing back took time when a quality issue arose.
Technical stack (2024–2025) PLM — Order-based structure: the order itself is the base unit of design and production
MES — process-result collection, LOT-level history logging, real-time equipment monitoring
FLOW — order → design → production → shipment as a single data flow on standard tables
Result [CUSTOMER RESULT]100% automated LOT traceability — tracing back a quality issue went from manual cross-checking to a lookup. Real-time equipment monitoring established.
CASE C-04 · MES → MANUFACTURING-AI PROGRAM

NKID — extending to AI
on the same standard

Die-casting industry mood imageINDUSTRY MOOD IMAGE
ProblemAluminum die casting — Phase 1 built an MES to establish a data foundation, but the next step, prediction and optimization, had to continue on the same standard rather than as a separate project.
Technical stack (Phase 1 2023.09–2024.09 · Phase 2 2025.10–2026.06) PHASE 1 — MES: accumulating process-result and quality data on standard tables
PHASE 2 — predictive maintenance (LSTM AE+CNN) · quality prediction (LSTM-CNN+Attention) · process optimization AI
INFRA — SaaS-model MES/FEMS demonstration (easing the initial burden for SMEs) · X-RAY 3D CT inspection integration
Why it mattersThis is where the difference from an SI that disappears once the build ends shows. The standard tables and ontology left from Phase 1 became the foundation for the Phase 2 AI expansion, and the customer moved to the next step on the same platform. The Phase 2 target figures are still at the planning stage, so they are not shown.

Where should our factory start?

In the diagnosis we look at your data structure and propose a starting point.

Request a Diagnosis