AI by / for Human, Health and Industry – AIby4

The four founders of i-Site NExT (University of Nantes, Centrale Nantes, INSERM and CHU) are working on a roadmap to structure AI for Nantes aligned with the objectives of NExT (Future Health & Industry). The shared vision encompasses two levels: 1) develop own research in interdisciplinary manner to address the value of IA in two applied sectors; 2) capitalize on own research with ecosystems and industries.
The joint scientific roadmap of the first level is articulated around the motto “AI made by Humans for Humans”. The design and evolution of this roadmap are notably supported by i-Site NExT funded clusters. NExT clusters are mostly "think tank" aiming to identify and structure research and education around key topics. Four existing funded NExt clusters (250K euros each) are connected to the AI roadmap: DELPHI : A DEep Learning Proposal for Health sciences & Innovation, GENDER: From gender studies in humanities and social sciences to emerging health and digital issues, FAME : Human Factors for Medical Technologies, IBD : Precision Medicine for Chronic Inflammatory Bowel Diseases.

The scientific roadmap combines 4 main research axis:
- 2 theoretical axes: AI by Human & AI for Human
- 2 applications: Future Health & Industry

These axes will leverage the existing local AI expertise, data-acquisition platforms, including secured biomedical data platform and clusters. Through this roadmap, we expect to develop a holistic approach to AI, to target the two applications domain combining local forces in:  
- AI methods: deep learning, Bayesian networks, topological data analysis, signal processing, computer vision, natural language processing
- Medical and Industrial applications: Genomics, medical imaging, clinical and care data, robotics, smart manufacturing, city and components
- Humanities: Gender bias, acceptability and social impact of AI, regulatory and GDPR, transparent governance of data an algorithms

From an implementation view point, two key points are targeted:
- the ambition to select and support at least a cohort of 20 interdisciplinary doctoral projects towards the development of novel methods that consider - the human aspect and societal impact of AI while answering to scientific questions in the two main application domains.
- the development of a joint training and scientific animation around the cohort, strengthening not only interaction and dissemination but also the structuring of AI for Nantes at scale considering many local labs and doctoral schools. 

Note that the current proposal, Alby4's aims at partially co fund this first level roadmap (15 PhD of the cohort and 100% of the training). Requested support from ANR being restricted to the UN cohort (15 Half-PhD). The other co-funding of the cohort is guaranteed by the NExT Initiative and Centrale Nantes and will be supplemented by funding to be obtained from other partners. The training funding is already fully consolidated.