Current Funded Projects

Reduce Hallucination in Large Language Models via Knowledge-Based Reasoning

ARC Discovery Project 2026–present Chief Investigator

This project aims to address the critical challenge of hallucination -- a phenomenon where generative AI models produce information that appears plausible but is factually incorrect -- with a focus on news fact-checking. The project expects to advance knowledge in detecting and mitigating hallucinations by developing innovative techniques for integrating external factual knowledge into AI models. Expected outcomes include a suite of innovative techniques to enhance AI models’ capability to reason and generate grounded information for complex fact-checking tasks. This should provide significant benefits, including improved reliability for generative AI systems and more effective action against misinformation at scale.

Trustworthy AI LLM Reasoning

AI4You(th): Advancing Generative AI for Personalized and Collaborative Youth Mental Health Care

Wellcome Trust 2026–present Collaborator Core Project Team

The global youth mental health (YMH) crisis demands scalable, effective solutions, yet current care models remain insufficiently engaging and effective. Critical barriers include limited personalisation, reduced youth agency in collaborative decision-making, and poor care continuity. Generative AI has the potential to address these challenges. However, key limitations prevent clinical translation: LLMs cannot reason longitudinally over evolving clinical contexts; lack frameworks for ensuring safety, youth agency, and evidence-informed decision-making; and operate as standalone tools disconnected from face-to-face care. This project will address these limitations through SensAI, a multi-stakeholder AI collaborator that supports young people with depression and anxiety and their clinicians to enhance personalisation, agency, and continuity in YMH.

Health AI Trustworthy AI AI for Science

Artificial Intelligence in Collaborative Robotics for Advanced Manufacturing Factories

Global Science and Technology Diplomacy Fund (GSTDF) 2026–present Chief Investigator

This project explores an AI-driven framework for integrating sensing, communication, and computation in swarm robotic systems within Industry 4.0 and emerging Industry 5.0 manufacturing environments. As production systems become more decentralised and adaptive, robotic swarms offer scalable and resilient alternatives to traditional automation, enabling collaborative assembly, autonomous inspection, and dynamic workflow reconfiguration. To realise their full potential, swarm robotics must tightly integrate three core capabilities: advanced sensing, industrial wireless communication, and AI algorithm-hardware co-design. Sensors provide essential environmental perception and feedback; industrial wireless networks ensure reliable, low-latency data exchange; and co-designed AI algorithm-hardware accelerators enable energy-efficient perception and decision-making at the edge.

Collaborative Robotics Advanced Manufacturing

Neuromorphic Module for Covert Communications

AEA Ignite 2026–present Chief Investigator

This project aims to develop a new type of underwater communication technology. Traditional underwater communication methods such as sonar (acoustic communication) are slow, easily detected, and not secure. Radio waves do not travel well through seawater, while surfacing to collect or receive data from satellites can put underwater vehicles at risk. Our solution uses a specialised optical receiver that captures light-based signals from satellites, even through rough sea surfaces. It combines tiny lenses or metasurfaces with a special type of sensor that mimics how the human eye processes motion and changes in light. This allows the system to detect very faint and rapidly changing light signals in a way that is fast, energy-efficient, and difficult to detect. The project will deliver a working prototype tested in wave tanks that simulate real ocean conditions. It will improve how data is sent to and received by underwater systems, making environmental monitoring and offshore industries more secure, efficient, and autonomous.

AI for Telecommunications Neuromorphic Computing

Past projects

Biochemical Text Mining for Advancing Chemical and Pharmaceutical Knowledge

ARC Linkage Project 2018–2022 Postdoctoral Researcher ChEMU Shared-Task Lead

This project developed natural language processing methods to find, extract, and structure complex chemical reaction information from scientific literature and patents, helping transform biochemical research into actionable knowledge for drug discovery. As a postdoctoral researcher on the project, I led the ChEMU shared-task series, coordinating international evaluation campaigns on chemical named-entity recognition, reaction-event extraction, and reference resolution. The shared tasks released benchmark datasets and enabled research teams worldwide to develop and compare methods for chemical patent information extraction.

Biomedical NLP Chemical Information Extraction AI for Science