Dundalk IT researchers at the cutting edge of Artificial Intelligence

Dundalk Institute of Technology (DkIT) have reported that multiple members of their Regulated Software Research Centre (RSRC) presented their research at the 31st Irish Conference on Artificial Intelligence and Cognitive Science (AICS) at the Atlantic Technological University in Letterkenny recently. 

The Regulated Software Research Centre (RSRC) in DkIT is a world leader in the development of methods and tools to assist medical device manufacturers to comply with the regulatory requirements related to the development of medical device software, whether embedded within the device, standalone, or as part of a networked medical device.

The research in the centre is conducted in collaboration with industry partners, national agencies, the international standards community and other research centres. The research team are active in the development of international standards and technical reports related to medical device software development and IT networks.

The PhD candidates Karla Cepeda, Buddhika Jayaneththi and Maliheh Heidarpour each presented their work on Data Security Challenges in AI-enabled Medical Device Software, A review of the Artificial Intelligence Act Proposal and the Medical Device Regulation and Pre-processing Techniques to mitigate Against Algorithmic Bias. 

In addition, Dr Róisín Loughran, from Dundalk and manager of the RSRC, was an invited keynote speaker at this conference and delivered a presentation on 'Regularity and Ethical Concerns for AI in Critical and Non-Critical Domains'.

Dr Loughran spoke about Artificial Intelligence (AI) which is an extremely powerful tool, with great potential benefits, she said “but we must always be careful to consider the ethical, regulatory and social challenges that misuse, or careless use, of such technologies could pose, both in critical domains such as healthcare but also in non-critical, creative domains”.

In this talk Dr Loughran introduced the Regulated Software Research Centre and the work they are undertaking to help developers, users and policy makers regulate AI in the safety-critical domain of medical device software.

She added: “I specifically look at detrimental biases in the world around us and how AI can make these biases worse if we are not careful. Standards, policies and guidelines can be proposed to help alleviate this, but ultimately, we must be cognizant of our own unconscious biases to ensure we do no harm when employing new technologies such as AI”. 

Dr Róisín Loughran also brought the participants through a view of generative systems in creative domains such as music and art and to consider what challenges and opportunities can arise from using AI for these purposes. A musical AI system is highly unlikely to cause physical harm and can offer opportunities for music generation to those who may not have the ability or access to such amenities. But creative generative systems could also reduce opportunities for upcoming artists and can cause significant unrest for many who work in these fields.

She further added “regardless of the application domain, we must always consider the implications of introducing AI into any system and the impact that it will have on those around us”.