← Cases
South KoreaEnergy and utilities2018 – ongoing

Samchully

Deep-learning recognition of gas-meter photographs

Samchully put AI gas-meter photo recognition into field use in February 2020

WorkingCompany's own claimOfficial ↗

Problem

Field staff read gas meters visually during service visits. Manual interpretation created opportunities for reading errors.

Approach

Samchully began working with an AI specialist in 2018 to develop a meter-photo recognition algorithm. After approximately 2 years of deep-learning development, it introduced the system for meter replacements in February 2020. Use extended to supply suspension, moving-related service and regular readings.

Results

No published figures

Takeaway

Samchully reused photo recognition across several field-service tasks. It reported fewer opportunities for manual-reading errors without quantifying the reduction.

Categories

Business function
Production and quality
Innovation type
Process innovation
AI technology
Computer vision
Evidence
Company's own claim

Sources

  1. 삼천리 70주년: Chapter 4, Section 5
    Samchully2025Company publication

Source summaries

[1]Samchully2025

Samchully's corporate history dated development to 2018 and operational deployment to February 2020. It described expansion across field-service work and claimed better efficiency and fewer manual-reading errors, without numerical results.

Full details are in the original sources.

Added 1 Nov 2026 · Updated 1 Nov 2026
Open case page ↗