ESR16 - Irakli Pkhakadze

ESR16 has completed a study exploring how synthetic data and Digital Twin technologies can work together to improve AAL systems, particularly in the areas of privacy protection, personalisation, and predictive accuracy. AAL technologies are essential for helping older adults live independently for longer, but they rely on collecting sensitive health and behavioural data, raising privacy concerns. The research highlights the value of synthetic data as a privacy-preserving method that still allows for useful analysis, especially when dealing with rare events. When synthetic data is embedded into DT frameworks, it supports real-time monitoring and scenario simulations, enhancing the responsiveness of AAL environments.

To ensure fairness and trust, the study uses bias-mitigation techniques and Explainable AI tools to improve transparency and reduce discrimination in AI-driven decisions. It also stresses the importance of compliance with the GDPR and the integration of Privacy by Design (PbD) principles from the early stages of development. Using both quantitative and qualitative methods, including the analysis of generative models like GANs and VAEs, the research finds that proper anonymisation and pseudonymisation reduce risks while enabling lawful data use. The study concludes that DTs, when combined with synthetic data and privacy-focused design, offer a promising solution for ethically and legally sound AAL systems. It calls for interdisciplinary collaboration to align technical and legal approaches and to develop robust, scalable frameworks that support both innovation and privacy.