References

Change Floor Operators Can Feel

The results the better's solutions have delivered at real industrial sites.

CASE 01 · Automotive Exterior-Parts Painting

Jinsung S&P

We built an On-Device AI machine-vision counting solution for Jinsung S&P, an automotive exterior-parts painting company, delivering stable, fast performance even in tight and complex spaces.

2022 AI Voucher Support Program · final evaluation completed

Jinsung S&P UV painting line — automotive exterior-parts painting process — Part of the Jinsung S&P UV painting line

OBJECTIVES

Objectives

01 Prevent human error and productivity loss caused by manual inspection
02 Apply a high-speed, low-power, low-cost solution based on Edge Device and Vision AI

OUTCOMES

Outcomes

Automated counting minimized workload and improved productivity
Made-to-order production cut warehousing costs
Buying materials in exact quantities to match output reduced purchasing costs

KEY FEATURES

Key Features

Counting unaffected by environment or object
Edge Device that maximizes space efficiency
Short installation time, minimal requirements

APPLICATION PHOTOS

In Real-World Use

A self-designed Edge Device installed compactly in a tight space — Space-efficient Edge Device

A compact camera mounted on the painting line — minimal installation requirements — Mini camera installation

Vision AI output detecting and counting automotive exterior parts in real time — Counting screen (object detection)

PERFORMANCE METRICS

Verified Performance Metrics

58.45 mAP

Object detection accuracy

Exceeded the 35 mAP target

29.8 FPS

Edge inference speed

Exceeded the 20 FPS target

80 types+

Real-time automotive-parts detection

Stable counting regardless of environment or object

5 %

Cost reduction achieved

Exact-quantity material purchasing · made-to-order production

CERTIFICATION

Official test report issued by KCL (Korea Conformity Laboratories)

Third-party certified verification of object detection accuracy and inference speed metrics

CASE 02 · Food Manufacturing Smart Factory

Sangshin Food

A multi-angle smart-factory deployment across the key processes of food manufacturer Sangshin Food, integrating multiple modules — temperature-deviation monitoring, production statistics, line-status checks, dashboards and vision recognition — by linking our own delivered hardware and software.

OBJECTIVES

Objectives

01 Production Built smart core-temperature data collection and alarm IoT
02 Post-processing Built an integrated system for smart sealing + automatic date-stamping + vision inspection
03 Shipping Built a real-time temperature monitoring/alarm system for finished-goods freezer storage
04 Demand ERP: automated mobile ordering and integrated management

OUTCOMES

Outcomes

Digitized handwritten HACCP product-temperature reports across all lines
Strengthened expiry-date and labeling imprints through Vision Recognition
Real-time digitization of temperature records previously dependent on paper logs

APPLICATION PHOTOS