Secure, Trustworthy and Grounded Digital Production Assistant (STIG)

Purpose and Objectives

Project STIG develops the next generation of AI-based decision support for manufacturing. By combining production data, domain expertise and advanced AI methods, the project aims to deliver trustworthy, secure and traceable recommendations that help companies understand, analyse and solve production challenges. The objective is to demonstrate the solution in real industrial environments and strengthen industry's ability to adopt AI in an effective and responsible way. The project builds on the results, methods and industrial collaborations established in the previous SIFT project to develop the next generation of trustworthy AI-based decision support for manufacturing.

Challenge

Manufacturing companies collect vast amounts of data but often struggle to quickly identify the root causes of production disturbances, quality issues and inefficiencies. At the same time, many AI solutions are difficult to understand and trust in business-critical production environments. To create real industrial value, AI solutions must be transparent, secure and grounded in both data and human expertise. It is a major challenge for manufacturing companies to manage and create value from large, fragmented datasets.

Implementation and Partners

RISE leads the project and develops methodologies and AI solutions together with Chalmers. Nexans and Profilgruppen contribute industrial use cases, production data, and testing environments. IFM Electronic and Good Solutions provide solutions for data collection, system integration, and the digitalization of production environments.

The STIG project will be demonstrated at Nexans and Profilgruppen with the objective of reaching Technology Readiness Level (TRL) 6 and contributing to increased digital resilience, safer AI implementation, and enhanced competitiveness for Swedish industry.

Together, the project partners ensure that the project results are both scientifically robust and practically applicable in industrial settings.

Solution

The STIG project develops a trustworthy, secure, and traceable AI-based digital assistant that combines process data, knowledge graphs, production ontologies, and industrial domain expertise. The solution enables users to ask questions in natural language and receive explainable recommendations based on both operational data and known cause-and-effect relationships in production. The technology can be applied to tasks such as troubleshooting, quality improvement, production optimization, and maintenance.

The project builds on the results, methodologies, and industrial collaborations established in the SIFT project (Vinnova Grant No. 2024-02480). SIFT developed an application for searching and visualizing time-series data from manufacturing processes and linking this data to relevant metadata such as quality outcomes, production disturbances, product variants, materials, and order numbers.

In the STIG project, this data-driven foundation is further enhanced by integrating domain expertise as an additional and highly valuable source of knowledge. This enables the AI assistant to combine data-driven insights with human expertise to provide more reliable, contextualized, and actionable decision support for industrial users.

Impact

The project is expected to contribute to increased productivity, improved product quality, faster problem resolution and reduced material waste. By strengthening companies' ability to adopt trustworthy AI, the project supports a more competitive, resilient and sustainable manufacturing sector. The project contributes particularly to UN Sustainable Development Goal 9: Industry, Innovation and Infrastructure and Goal 12: Responsible Consumption and Production through improved resource efficiency and reduced waste.

Funded by Vinnova

The project is funded by Vinnova through the Advanced Digitalisation programme under the call "Digital Resilience, AI and Cybersecurity for a Competitive Industry".