공급망 리스크 조기 경보에 대한 베이지안 네트워크 모델 적용 가능성: 반도체 산업을 중심으로 (The Applicability of Bayesian Networks Models for Supply Chain Risk Early Warning: Focus on the Semiconductor Industry)

서지정보

· 발행기관 : 한국유통경영학회
· 수록지 정보 : 유통경영학회지 / 26권 / 5호 / 119 ~ 133페이지
· 저자명 : 신성호, 신수용, 김동환, 배성훈

초록

Purpose: This study analyze the current status of the import supply chain of materials, parts, and equipment required for semiconductor production by air and sea transport, and to identify ways to secure a stable import supply chain by comprehensively identifying the current status, stability, vulnerabilities, and risks of domestic and foreign supply sources using Bayesian Networks models.
Research design, data, and methodology: Using TRASS (Trade Statistics Service) data, we collected and classified Harmonized System (HS) codes for semiconductor subheadings and extracted import origin point data (2015, 2020, and the first half of 2022; 3 years). Using the origin point analysis method and Bayesian Networks, the supply chain status of selected equipment by country, item, and mode of transport was identified. We presented examples of the application of Bayesian Networks.
Results: It was confirmed that Bayesian networks can be used to map the supply chain of imports from a country from a supply chain perspective with HS code ten units by transport mode and port or airport and to prepare early warning and countermeasures for the import chain of the item in case of supply chain risks. Thus, Bayesian Networks can build supply chain mapping, that can be extended to early warning.
Implications: Supply chain mapping provides a view of each link used to more quickly detect the risk of supply chain disruption from actual raw materials to production facilities, parts, and intermediate materials. This study is significant as it presents a model that indicates warnings for items that increase supply chain risk when detecting risk by country through Bayesian Networks. It also suggests the possibility of using this methodology in supply chain risk in the future by applying Bayesian Networks.

영어초록

Purpose: This study analyze the current status of the import supply chain of materials, parts, and equipment required for semiconductor production by air and sea transport, and to identify ways to secure a stable import supply chain by comprehensively identifying the current status, stability, vulnerabilities, and risks of domestic and foreign supply sources using Bayesian Networks models.
Research design, data, and methodology: Using TRASS (Trade Statistics Service) data, we collected and classified Harmonized System (HS) codes for semiconductor subheadings and extracted import origin point data (2015, 2020, and the first half of 2022; 3 years). Using the origin point analysis method and Bayesian Networks, the supply chain status of selected equipment by country, item, and mode of transport was identified. We presented examples of the application of Bayesian Networks.
Results: It was confirmed that Bayesian networks can be used to map the supply chain of imports from a country from a supply chain perspective with HS code ten units by transport mode and port or airport and to prepare early warning and countermeasures for the import chain of the item in case of supply chain risks. Thus, Bayesian Networks can build supply chain mapping, that can be extended to early warning.
Implications: Supply chain mapping provides a view of each link used to more quickly detect the risk of supply chain disruption from actual raw materials to production facilities, parts, and intermediate materials. This study is significant as it presents a model that indicates warnings for items that increase supply chain risk when detecting risk by country through Bayesian Networks. It also suggests the possibility of using this methodology in supply chain risk in the future by applying Bayesian Networks.

참고자료

· 없음