Building an ethical framework for human-centred AI in smart manufacturing
Who I helped
In the MAS4AI project (Multi-Agent Systems for Pervasive AI for assisting Humans in Modular Production), funded under Horizon 2020, I co-created and led the development of the project’s ethical foundation. As an ethics and legal expert, I authored the MAS4AI Ethics Framework, the Ethical AI Risk Toolkit and contributed to the final validation of the Ethics Framework across all industrial pilot sites.
The project aimed to develop multi-agent AI systems that support human decision-making in manufacturing, combining autonomy, adaptability and safety with a strong ethical foundation.
Project’s webpage: https://mas4ai.eu/
Context and challenge
MAS4AI sought to integrate AI-driven agents into real production environments, enabling collaboration between humans and intelligent systems.
Key ethical challenges included:
- protection of personal and sensor-based data collected in smart factories,
- risk of excessive monitoring and impact on workers’ rights,
- ensuring explainability and transparency in automated decisions,
- establishing clear accountability and human oversight mechanisms,
- integrating legal requirements into industrial research and innovation.
What I did
- authored the MAS4AI Ethics Framework, defining ethical principles, roles and assessment methods for AI development in smart manufacturing,
- developed the Ethical AI Risk Toolkit, a practical instrument to identify, prioritise and mitigate ethical risks throughout the project lifecycle,
- led the final validation of the Ethics Framework in the SmartFactory-KL testbed and five industrial pilot environments, confirming its applicability to AI-based multi-agent systems,
- introduced the ethics-by-design approach, embedding ethical reflection into technical development, testing and decision-making processes,
- delivered workshops and lectures on AI regulation, GDPR, and Trustworthy AI, guiding partners in aligning their industrial pilots with EU legal and ethical standards,
- prepared of the final ethics deliverable (Ethics Framework Validation Report), confirming that all pilots successfully implemented the framework.
Results and value
- the project validated a comprehensive and operational Ethics Framework applicable to AI-based agents in industrial environments,
- the consortium implemented ethical risk management procedures across all pilot sites, ensuring proportional and consistent application of ethical standards,
- the Ethical AI Risk Toolkit became a reference for partners to assess and manage risks linked to data, fairness and human oversight,
- the validation process demonstrated that ethics can be measured, monitored and improved in AI-driven industrial systems,
- MAS4AI successfully delivered all planned ethics outcomes, meeting EU expectations and setting a model for future Industry 4.0 projects.
Summary / invitation
In MAS4AI, I helped design, implement and validate a complete ethics governance system for industrial AI – from framework to practical tools and real-world verification.
If your project involves AI for manufacturing or industrial automation, I can support you in building and applying ethical frameworks that ensure compliance, trust and sustainable innovation.
