How our research works
A2I2 tackles real-world problems using methods grounded in modern machine learning, deep learning and software engineering practices to make discoveries to advance knowledge and build systems that benefit society.
By exploring new frontiers in AI and implementing safe and effective AI solutions, the Institute transforms industries and improves people's lives around the world.
Our research themes
A2I2 is committed to pushing the boundaries of AI. The institute is a top Australian research body focusing on Human-centred AI, and our researchers regularly publish in top international AI conferences (NeurIPS, ICML, ICLR, AAAI, IJCAI, CVPR, ICSE) and Q1 journals.
Our research themes cover:
- Bayesian optimisation
- Deep learning and beyond
- Reinforcement learning
- Responsible AI, assurance and robustness
- AI for software engineering, and software engineering for AI
A2I2 are pioneering methods and approaches for AI and machine learning problems with high potential impact.
Help to positively impact our world
As a PhD student at A2I2 you will learn alongside the sharpest minds in the field of AI, working in world-class facilities at the beautiful campuses in Burwood in the heart of Melbourne and Waurn Ponds in the vibrant regional city of Geelong.
In the future, AI will become ubiquitous. With this vision in mind, we aim to build AI that is useful to society, reliable, and aligned to our core human needs and values.
— Alfred Deakin Professor Svetha Venkatesh
Alfred Deakin Professor Svetha Venkatesh
ARC Laureate Fellow, Co-Director, Applied Artificial Intelligence Institute (A2I2)
Alfred Deakin Prof Venkatesh’s work is focused on developing cutting-edge solutions to problems of societal importance in critical areas like health, security, manufacturing and aged care. Her work has resulted in the ability to detect potential security threats in large data sets; a health analytics program that helps doctors predict suicide risk; and an early intervention system for children with autism.
Alfred Deakin Professor Kon Mouzakis
Co-Director, Applied Artificial Intelligence Institute (A2I2)
The work by Alfred Deakin Prof Mouzakis involves studying how the implementation of artificial intelligence (AI) into complex systems can improve our understanding of some of the more difficult problems we all face in this challenging world.
Associate Professor Sunil Gupta
Head of AI, Optimisation and Materials Discovery, Applied Artificial Intelligence Institute (A2I2)
A/Prof Gupta's research interests include development of theory and practice for sample-efficient AI and machine learning such as Bayesian optimisation, multi-arm bandits, reinforcement learning and transfer learning. He has focused on applications in healthcare, computer vision, advanced manufacturing and materials discovery.
Professor Rajesh Vasa
Head of Translational Research, Applied Artificial Intelligence Institute (A2I2)
Prof Vasa's research focus is in the area of robust AI-infused technology that assists decision making in real-world contexts. His work has led to the creation of two start-ups, and his research outcomes have delivered consumer facing products and commercial systems that are in production.
Talk with Ted
This industry-leading digital training avatar educates nurses on delivering the best standard of care for people living with dementia. Our research trial highlighted that carers who train using Talk with Ted develop a better understanding of the behaviour of people living with dementia and how poor communication can trigger changed behaviour.
Early Cerebral Palsy Screening
Early detection matters in neurodevelopmental disorders. We are developing and validating new methods that address the challenges associated with early-stage screening, and aim to make the process faster, more convenient, low-cost, and ultimately more accessible to the broader community.
Using AI-based adaptive trial methods, this world-first study aims to discover the most effective treatments for mental illnesses and to assist clinicians and health services to make better decisions through precision medicine.
Vital Monitoring System
The Vital Monitoring System (VMS) is a clinically proven risk management technology that identifies and predicts deterioration of respiratory illness in patients. It has remotely monitored over 1400 COVID-positive patients, relieving the strain on the Victorian hospital system.
Our state-of-the-art facilities help us to conduct transformative work.
AI computing cluster
Optimised to support AI research, the cluster includes a range of powerful servers that meet the demands of modern research and development needs. The cluster boosts our experimentation capabilities with the ability to run complex AI models.
C2 Command and Control Infrastructure
A laboratory designed to support research and development of applications and processes used by mission commanders during mission planning and execution. Applications include support to Emergency Services, Health (Ambulance and Hospital) and the Military.
Our simulation facilities are dedicated to supporting the creation and visualisation of 3D environments in Augmented and Virtual Reality. With our equipment suite, we are able to design, develop and test VR and AR simulations for industry partners.
A2I2 collaborates with a broad range of research and innovation partners to produce and deliver high impact solutions.
Our areas of application span health and AI, education technologies, defence and cyber security, advanced manufacturing and more. Recent examples of our collaboration outcomes include reducing errors in hospital trauma bays to save lives, tackling Australia's aged care quality crisis, and contributing to Australia's largest recycling and clean energy advanced manufacturing ecosystem.