Other international projects
Department of Environmental Sciences, Informatics and Statistics

The HEALTH-AI project aims to bridge the gap between healthcare and Artificial Intelligence (AI) by providing professionals and students with the knowledge, skills, and ethical awareness necessary for the safe and effective integration of AI into clinical practice. It aims to improve AI skills through an adaptive AI curriculum, strengthen interdisciplinary collaboration and knowledge sharing between informatics scientists and healthcare professionals through a co-creation platform, and promote responsible AI adoption through appropriate ethical guidelines.

The EcoDigify project aims to build university-level training courses of an interdisciplinary nature to promote sustainability in the use and development of digital technologies. This will be achieved by providing teachers with materials and tools on sustainable digitalisation and providing students with the knowledge and skills to create sustainable digital solutions with the support of experts and researchers working in the field of sustainability.

ATLAS aims to bring about positive change for local authorities and stakeholders in the assessment and management of the relationship between green spaces and historic buildings. The project proposes an approach to risk analysis with the aim of developing innovative tools that facilitate the assessment of the multiple hazards, vulnerabilities and risks associated with the conservation of heritage and green spaces in historic cities, while ensuring careful use of public funds. In addition, protocols will be established to promote the implementation of sustainable adaptation and mitigation strategies based on integrated and participatory management of green spaces and heritage resources.

AWS IAM access control policies specify what AWS cloud resources are accessible by what identities. AWS SDKs allow developers to programmatically access these resources through applications that can be run by different identities. However, what resources are accessed and how depend on the (mostly string) values computed by the application. Therefore, forecasting what access control policy is required by a given application is rather complex. In this project, we apply abstract interpretation techniques to over approximate the string values computed by these applications, and then infer, validate or IAM the access control policies.