10 Year Anniversary

Colloquium: Envision the Next 10 Years

• Demo I: Neuro-symbolic Artificial Intelligence for Asset Management & Industry (Dr. Ploennigs)
  Automating Artificial Intelligence techniques for IoT applications is key to scaling the exponential growth of data created from the widespread adoption of IoT to monitor factories, buildings, and cities. IBM Researcher in Dublin are using neuro-symbolic AI approaches that combine domain knowledge in the form of semantic knowledge graphs with machine learning and optimization approaches to enable data scientists and domain experts alike to automate the creation of Digital Twins - a virtual model that reflects a physical object. Dr. Ploennigs will explain and demonstrate how this works in practical applications such as robots and electric race cars.
• Demo II: AI & Social Determinants of Health (Dr. Bettencourt-Silva)
  This demo will showcase how Artificial Intelligence, Bayesian Networks, Natural Language Processing and Knowledge Graphs can support government agencies in understanding the social program levers such as housing, transportation, nutrition or financial security that drive positive health and well-being outcomes at a population level. IBM Researchers at the Dublin Lab have created interconnected AI models to capture the influence of Social Determinants of Health (SDoH) on health outcomes. By linking typically siloed datasets and heterogeneous textual sources, these AI models enables better predictions that have the potential to improve access and equality for our citizens, drive positive health and well-being outcomes, and relieve resource burdens.
• Demo III: Blending Applied Math and AI for Real-world Insight (Dr. Sergiz Zhuk)
  We are surrounded by examples of dynamical, real-world phenomena that are understood by science but where applying AI technologies remains a challenge. Could a synergy between the mathematics underpinning bio-physical processes and AI provide useful new insights? For example, the technique of adding tracers, such as a dye, to trace fluid movement is used widely in areas such as climate science, ecology, civil engineering and medicine. The underlying process of dye dispersal is well understood as a mathematical problem, allowing researchers to model how pollutants move or to detect leaks. Dublin researchers are using a blend of applied mathematics and AI to generate insights by spotting subtle differences in the way fluids move through materials. In this talk, Dr. Zhuk will showcase a first-of-a-kind prototype, which was developed for the Mater Misericordiae Hospital at the University College Dublin and the Royal College of Surgeons Ireland, and supported by Irish Disruptive Technology Innovation Fund. His demo will illustrate how computer vision, optimization and classification algorithms can work together to extract insights from fluid movement in videos, including multispectral endoscopic videos, in nearly real time.

Closing