MCS — Material Control System
2022.02 – present(4 yrs 9 mos)OSAT C, IncheonLead · team of 3
The transport execution system: breaks each transport request into equipment-level commands for AGVs, AMRs, stockers and buffer stations, and tracks every move end to end.
- AGVs & AMRs connected
- 50+
- stockers integrated
- 10+
- equipment comms standard
- HSMS-SS
- Splits RTD commands into equipment-level steps, tracks them and reports failures at once
- Covers the full command lifecycle, from transfer, cancel, abort and replace to state sync on restart
- Puts ACS, FMS and 10+ stockers behind one interface, extended to buffer stations and magazine handling
RTD — Real-Time Dispatcher
2022.04 – present(4 yrs 7 mos)OSAT C, IncheonLead · team of 3
The transport decision system: applies per-process rules to decide in real time what to move, from where, to where and by which means of transport.
- back-end processes
- 10+
- dispatching rules
- ~50p
- transport transactions / day
- 2,000+
- dispatch decision
- < 1s
- Defined dispatching rules with the process engineering teams and tuned them on site
CIM — Production System Integration
2022.11 – present(4 yrs)OSAT C, IncheonLead · team of 3
The integration gateway between the customer’s production system and the transport systems: a single point of contact that takes production requests and connects them to MCS, RTD and the rest of the suite.
- The customer’s first logistics automation — built every IT system for the process turnkey with MCS, RTD and CIM
Vision-Based AGV Lane Detection
2021.04 – 2024.02(2 yrs 11 mos)Packaging & Test S, CheonanSole developer
A vision guidance system for heavy-load AGVs that follows the driving path with cameras and an AI model instead of conventional sensors. Deployed on production-line vehicles and still running today.
- detection accuracy
- 99%+
- per tick
- 40ms
- vehicles deployed
- 50+
- in production
- 4 yrs
- Minimized driving errors compared with the sensor-based system
- Selected as one of the customer’s internal innovation items for 2023
- AI & Vision
- Python
- deep learning
- edge AI inference
- image processing
360° Vision-Based Vehicle Collision Prevention
2026.03 – present(8 mos)Lead · sole developer
A vision safety system that fuses stereo camera feeds around a vehicle and judges collision state in 3D monitoring zones in real time.
- coverage
- 360°
- cameras processed
- 4–8
- decision rate
- 30fps
- vehicles deployed
- 10+
- Processes multiple camera feeds in real time on edge GPUs and judges collision state per monitoring zone
- Operators manage monitoring setups and vehicles select the one that fits each task
- Web UI for setup and monitoring, with APIs for vehicles and external systems
- Vision
- C++
- GPU acceleration
- deep learning inference
- depth cameras
Logistics Equipment Management System
2024.01 – 2025.09(1 yr 9 mos)Packaging & Test S, CheonanLead · team of 2
Collects IoT sensor, driving and transport data from every logistics system in the field and bundles on-site service tools, raising automated production efficiency.
- AGVs & AMRs connected
- 100+
- Automatic program updates, SOP, work log and maintenance history, digital twin, alarm analysis and automatic reports
- Rolled out to the customer’s overseas plant
- Third layer on the same fleet, after driving control and vision guidance
Automated Battery Change System for AGVs
2025.01 – 2025.12(1 yr)Packaging & Test S, CheonanSole developer
Swaps the batteries of heavy-load AGVs automatically in a stocker-style unit and manages their charging.
- Automatic battery loading, unloading and charge management
- In production and slated as the customer’s back-end line standard
SEMI E5/E37 (SECS-II/HSMS) Communication Library
2026.07 – present(4 mos)Sole developer
Defined a canonical specification from SEMI E37 (HSMS) and E5 (SECS-II) and wrote two implementations, Java and .NET, from the ground up with no external dependencies. Deployed in field systems.
- implementations (Java · .NET)
- 2
- external dependencies
- 0
Team Platform — Automation Backend Engine · Web UI Library
2023.12 – present(2 yrs 11 mos)Lead
The backend engine, web UI library and development tooling used across the logistics software team.
- Automation backend engine — a general-purpose .NET engine for factory automation, used by every developer on the team
- AI coding agent templates — linked to the issue tracker and used across the team
- Web UI library (sole developer) — the shared React library behind every web stack on the team; forms, a high-level data grid, workspace tabs and theming with no UI framework dependency
- Automation engine
- C#
- .NET
- industrial communication
- equipment control
