Periodic Reporting for period 2 - SC4EU (True demand-driven semiconductor supply chains for Europe)
Período documentado: 2024-12-01 hasta 2025-11-30
SC4EU, a collaborative Innovation Action aims at strengthening European digital sovereignty by mitigation of the chip shortage through reduction of bullwhip effect in the semiconductor industry and supply chains containing semiconductors. This will be reached via a “truer”-demand signal gained from an anonymous MPC (Multi Party Computation) survey on coarse granularity which will be broken down via AI methods to fine granularities following the Semantic Web-based Digital Reference structure.
The ambition of SC4EU consortium is to obtain high-quality, reliable data for semiconductor demand forecasting. In the solution proposed by SC4EU data should be gathered via an anonymous survey based on Multi-Party Computation technology. Anonymity and security of data flow will encourage business partners to share their true demand data. Then, the gathered data will be mapped onto ontologies (semantic representations of the semiconductor industry) and processed with AI tools for demand breakdown of fine granularity.
SC4EU will result in a better bullwhip reduced demand forecast on all levels of the semiconductor demand and of demands for supply chains containing semiconductors. SC4EU is a pan-European initiative to boost industrial competitiveness through interdisciplinary innovations, establishing sustainable ECS value chains and therefore contribute to European digital sovereignty addressing urgent issues like security of supply, monitoring and crisis response, and chip shortage.
The ambition of SC4EU consortium is to obtain high-quality, reliable data for semiconductor demand forecasting. In the solution proposed by SC4EU data should be gathered via an anonymous survey based on Multi-Party Computation technology. Anonymity and security of data flow will encourage business partners to share their true demand data. Then, the gathered data will be mapped onto ontologies (semantic representations of the semiconductor industry) and processed with AI tools for demand breakdown of fine granularity.
SC4EU will result in a better bullwhip reduced demand forecast on all levels of the semiconductor demand and of demands for supply chains containing semiconductors. SC4EU is a pan-European initiative to boost industrial competitiveness through interdisciplinary innovations, establishing sustainable ECS value chains and therefore contribute to European digital sovereignty addressing urgent issues like security of supply, monitoring and crisis response, and chip shortage.
During period 2, SC4EU continued active development and coordination towards the Digital Reference ecosystem and the asso-ciated user administration processes. IFAG led and moderated regular meetings, ensuring that all bene-ficiaries’ efforts were effectively aligned and integrated across the work package. Progress was made in expanding the ecosystem's technical capabilities and collaboration between partners, although the complexity of several tasks led to necessary extensions and adjustments to initial planning. The OCP was significantly expanded with new and enhanced features. The vocabulary development environment now includes the Ontology Development Tool (OnDeT), which visualises the evolution of project terminology over time and is available for all ontologies registered in the OCP. The consensus tool was introduced to support controlled terminology maturation.
The planning ontology from the previous SC3 project was extended and refined towards a demand fore-casting ontology. New concepts, predicates, and agent activities in addition to the artifacts from the planning ontology were incorporated into the proposed forecasting ontology. The final version of the forecasting ontology was implemented in the Protégé tool. We also initiated to work on methods using the WSTS dataset and company-specific historical demand data from IFAG to forecast the IFAG demand based on the WSTS data. The System Dynamics (SD) model was further extended and refined. A model to predict lot completion times in wafer fabs was implemented and tested in the reporting period.
Significant results have been achieved concerning the specification and requirements for the survey use case. A substantial amount of dialogue has taken place among the project partners, focusing on clarifying the specifics of the survey questions. Given that different surveys are designed for each tier of the supply chain, the project partners have collaboratively discussed and tailored questions to meet the distinct needs of each tier. The development of the questions for the semiconductor manufacturers, the Tier-1 suppliers and the OEMs is completed and a successful first complete survey run has been conducted. Six survey types (SEMI, OEM, TIER 1, monthly/yearly) are now operational. Analysts can select survey templates that automatically configure the questions, MPC operations, and corresponding Excel import features connected to WP5 transformation files, ensuring privacy compliant data uploads.
In the first half of 2025 the “semi” survey questions were finalized within WP5. It was decided to cover the survey’s questions demand with the future sales & marketing forecasts, inventory, and order can-cellations, with a focus on technology nodes as dimension for the question. Next steps of WP5 is finish-ing the definition of the OEM questionnaire specification and shift the focus then to the Tier1 survey and the transformation graph between the semiconductor products. For the OEM survey the automotive market has been selected.
The planning ontology from the previous SC3 project was extended and refined towards a demand fore-casting ontology. New concepts, predicates, and agent activities in addition to the artifacts from the planning ontology were incorporated into the proposed forecasting ontology. The final version of the forecasting ontology was implemented in the Protégé tool. We also initiated to work on methods using the WSTS dataset and company-specific historical demand data from IFAG to forecast the IFAG demand based on the WSTS data. The System Dynamics (SD) model was further extended and refined. A model to predict lot completion times in wafer fabs was implemented and tested in the reporting period.
Significant results have been achieved concerning the specification and requirements for the survey use case. A substantial amount of dialogue has taken place among the project partners, focusing on clarifying the specifics of the survey questions. Given that different surveys are designed for each tier of the supply chain, the project partners have collaboratively discussed and tailored questions to meet the distinct needs of each tier. The development of the questions for the semiconductor manufacturers, the Tier-1 suppliers and the OEMs is completed and a successful first complete survey run has been conducted. Six survey types (SEMI, OEM, TIER 1, monthly/yearly) are now operational. Analysts can select survey templates that automatically configure the questions, MPC operations, and corresponding Excel import features connected to WP5 transformation files, ensuring privacy compliant data uploads.
In the first half of 2025 the “semi” survey questions were finalized within WP5. It was decided to cover the survey’s questions demand with the future sales & marketing forecasts, inventory, and order can-cellations, with a focus on technology nodes as dimension for the question. Next steps of WP5 is finish-ing the definition of the OEM questionnaire specification and shift the focus then to the Tier1 survey and the transformation graph between the semiconductor products. For the OEM survey the automotive market has been selected.
The SC4EU consortium’s approach for improving supply demand forecast in the semiconductor industry and mitigating bullwhip effect is based on:
- anonymous data sharing with use of Multi-Party Computation,
- semantic data processing to obtain high quality data of fine granularity,
- close cooperation of semiconductor companies and research institutes,
- focus on human factors,
- application of gamification and relate techniques for data sharing incentivisation.
SC4EU strives achieve the following outcome:
- Enforce Europe’s position in different fields of ECS, currently positioned behind the USA and China, in domains where autonomy, safety, precision are key elements – such as in automotive and other industries
- Ensure the European sovereignty in a booming market environment based on advanced ECS, by unlocking the potential of complex semiconductors operations, reducing their deployment costs and fostering their market introduction
- Influence standardisation through an already crystal-clear footprint
- Provide a credible route to delivering such European innovations to global markets. Also, nurture the users acceptance and adoption of innovative semiconductors and ECS as a part of the modern industrial and end user mix.
- Secure employment and create new jobs based on emerging services and applications on the basis of European ECS.
- anonymous data sharing with use of Multi-Party Computation,
- semantic data processing to obtain high quality data of fine granularity,
- close cooperation of semiconductor companies and research institutes,
- focus on human factors,
- application of gamification and relate techniques for data sharing incentivisation.
SC4EU strives achieve the following outcome:
- Enforce Europe’s position in different fields of ECS, currently positioned behind the USA and China, in domains where autonomy, safety, precision are key elements – such as in automotive and other industries
- Ensure the European sovereignty in a booming market environment based on advanced ECS, by unlocking the potential of complex semiconductors operations, reducing their deployment costs and fostering their market introduction
- Influence standardisation through an already crystal-clear footprint
- Provide a credible route to delivering such European innovations to global markets. Also, nurture the users acceptance and adoption of innovative semiconductors and ECS as a part of the modern industrial and end user mix.
- Secure employment and create new jobs based on emerging services and applications on the basis of European ECS.