Project Reference No. : C1073-25G
Project Title : Waste-derived iron salts enable integrated and sustainable urban wastewater management
Project Coordinator : Professor Zhiguo YUAN
University : City University of Hong Kong
Layman Summary
Wastewater treatment relies extensively on iron-based chemicals to control odour, remove pollutants, and enhance sludge treatment. At present, these chemicals are produced off-site and transported over long distances, resulting in high costs, significant carbon emissions, and potential occupational safety risks. This supply chain is susceptible to interruptions in other sectors.
This project proposes a sustainable and resilient alternative by producing iron salts locally through the valorisation of urban waste streams. Using renewable electricity, carbon dioxide captured from biogas and iron from municipal solid wastes will be converted into an iron compound that can be safely and effectively reused within wastewater treatment systems. In parallel, the process upgrades biogas into a cleaner and higher-value energy carrier, improving the overall efficiency of resource utilisation.
The project directly supports Hong Kong’s targets for carbon neutrality, waste reduction, and resilient urban infrastructure, while offering a scalable circular-economy model that can be adapted to cities worldwide.
Project Reference No. : C1077-25G
Project Title : Understanding & Fighting Scams: Psychological and Social Perspectives & Effective Interventions
Project Coordinator : Professor Shuhua ZHOU
University : City University of Hong Kong
Layman Summary
Scams cost the global economy over $1 trillion annually, yet countermeasures fail due to disciplinary fragmentation. This project pioneers the first unified “biopsychosocial” model by integrating AI, neuroscience, psychology, criminology, and communication through five interconnected tasks. The team will first employ AI-powered data analytics to construct a comprehensive typology of scam mechanisms. Simultaneously, the team will delineate the psychological processes and neural markers of victim vulnerability, alongside a criminological analysis of the societal structures facilitating fraud. These insights will drive evidence-based interventions that will be validated through randomized controlled trials and experimental studies. Finally, the practical deliverables of this project include an AI-driven detection system, an anti-scam gaming platform, a comprehensive data repository, and policy frameworks for scam inoculation and stakeholder-led prevention. This interdisciplinary and integrated approach represents a paradigm shift from fragmented knowledge to a scalable defense architecture, transforming how society protects itself against sophisticated scam operations that exploit human psychology and technological vulnerabilities.
Project Reference No. : C1134-25G
Project Title : Control of food intake and metabolism by a sensory cortex-periphery axis
Project Coordinator : Professor Chun Yue Geoffrey LAU
University : City University of Hong Kong
Layman Summary
While eating may appear to be a straightforward choice for humans, it is a complex decision for wild animals, influenced by hunger, bodily needs, and environmental risks. Olfactory cues are crucial for locating food, particularly when resources are limited. This project focuses on a less-explored brain region: the anterior piriform cortex (APC), the largest olfactory area responsible for processing smells. In particular, this research aims to uncover how the APC, a sensory cortex, coordinates with peripheral body signals to regulate food-seeking behavior, eating, and metabolism. A key hormone in this process is adiponectin, which influences energy expenditure and is associated with food intake. While adiponectin's effects on appetite centers in the brain are documented, its role in olfactory-related brain regions remains largely unexplored. This study employs advanced techniques, including multi-omic analysis, electrophysiology, and gene knockdown, to investigate how adiponectin affects APC neural circuits and behavior. The insights gained from this research could enhance our understanding of eating disorders and metabolic diseases such as diabetes, potentially paving the way for the development of novel therapeutic strategies.
Project Reference No. : C4025-25G
Project Title : Hong Kong's Window on the Universe - First Light
Project Coordinator : Professor Hua-bai LI
University : The Chinese University of Hong Kong
Layman Summary
The submillimeter band is the final frontier for wave detection and a tough spot for astronomy. Earth's atmosphere blocks most of these signals and adds heavy noise. While China has advanced in astronomy, it still lacks a submillimeter telescope, camera, or polarimeter. With ROGer on Greenland’s GLT, the team is taking a crucial first step.
It is an ambitious step: the polarimeter is aimed to be more efficient than traditional designs.
It is a creative step: the team has built a sky noise simulator to test ROGer realistically in the lab.
It is also a strategic step: before new GLT moves to its final site, the team will test ROGer on the world’s largest submillimeter telescope, the JCMT.
For Hong Kong, this is a landmark project in instrumentation. While Hong Kong promotes global collaboration, few projects are truly locally led and built from scratch. ROGer’s success will empower local astronomers and—more importantly—inspire students here to innovate, not just absorb existing knowledge.
Project Reference No. : C4085-25G
Project Title : Immunogenic Epitopes from the Non-coding Genome as Tumor-specific Antigens for Drug Development against Childhood Acute Myeloid Leukemia
Project Coordinator : Professor Kam-tong LEUNG
University : The Chinese University of Hong Kong
Layman Summary
Pediatric acute myeloid leukemia (AML) is a rare but aggressive malignancy with a complex genomic landscape but low mutational burden. Despite advances in chemotherapy and hematopoietic stem cell transplantation, 30-40% of patients will still progress to relapse or refractory diseases, with limited salvage options. Immunotherapy has revolutionized the treatment landscape for some types of hematologic malignancies, yet its application in AML remains limited due to the lack of tumor-specific antigens (TSAs). Recent studies in adult AML have identified cryptic TSAs derived from non-coding genomic regions, offering a promising antigen source for cancers with low mutational burden. Given pediatric and adult AML are genetically and biologically distinct, the TSAs identified in adult AML are unlikely representative of the pediatric population. Furthermore, due its rarity and the logistical challenges of acquiring sufficient samples to meet the stringent requirements of TSA discovery pipelines, no comparable initiative has been undertaken for pediatric AML. This project aims to construct the first comprehensive atlas of MHC-I and MHC-II–associated TSAs in pediatric AML using a dual-core proteogenomic strategy. The team will integrate mass spectrometry–based immunopeptidomics, transcriptomics, genomics, and ribosome profiling to build personalized peptide databases capturing canonical and non-canonical antigen sources. Identified TSAs will be evaluated for immunogenicity and therapeutic potential using computational prediction, ex vivo T cell assays, and single-cell TCR profiling. This study will yield a prioritized list of pediatric AML-specific TSAs with high tumor specificity and broad population coverage. The findings will lay the foundation for developing novel T cell-based immunotherapies tailored to pediatric AML, addressing a critical unmet need and advancing precision oncology for this unique disease entity.
Project Reference No. : C4097-25W
Project Title : A 5-Year Prospective Study of Multimodal Deep Phenotyping with Digital Measurements, Multi-Omics, and Neuroimaging Biomarkers in Prodromal and Early Stages of Alpha-Synucleinopathy Neurodegeneration
Project Coordinator : Professor Yun-kwok WING
University : The Chinese University of Hong Kong
Layman Summary
Neurodegenerative disorders like Parkinson’s disease and dementia with Lewy bodies (collectively known as α-synucleinopathies) are characterized by long 'prodromal' phases, a period before clinically significant symptoms and signs appear. This early phase represents a critical window for disease prevention, particularly given the lack of curative treatments for α-synucleinopathies. However, current clinical evaluation often misses the subtle and early signs of the diseases, and pathological sampling testing may be invasive (like spinal taps) and more importantly, not yet proven for tracking neurodegenerative progress. To solve this, the team is focusing on a specific sleep condition called isolated/idiopathic Rapid Eye Movement Sleep Behavior Disorder (iRBD) in which patients would physically ‘act out’ their dreams. This condition is currently the strongest known early warning sign and stage for α-synucleinopathies, with over 90% of people with iRBD eventually developing neurodegeneration within 15 years. In addition, the team’s family cohort data suggests that first-degree relatives of iRBD patients (RBD-FDR) are likely at an even earlier stage of the disease, harboring a spectrum of sleep features and neurodegenerative biomarkers.
Building on the team’s CRF (ref no.: C4044-21G), which has successfully established a prodromal and early cohort of α-synucleinopathy including healthy subjects, RBD-FDR, iRBD patients and those with early Parkinson’s disease, this renewal funding aims to extend the research for another 3 years. The extended project has three main goals: 1) Prospective digital monitoring: using the team’s validated smartphone app to detect subtle declines in motor and non-motor functions that standard clinical assessments might fail to capture accurately; 2) Integration of neuroimaging markers: combining brain scans with biological data (e.g., gut microbiota, inflammation markers, and proteomics) to better understand the underlying mechanisms of the disease and the varying progression trajectory across subjects; 3) Longitudinal tracking of multimodal data: by integrating clinical, digital, and biological features over time, the team aims to identify novel and reliable biomarkers and their dynamic and prospective changes for predicting neurodegenerative progression in high-risk population. This work will advance our understanding of disease mechanisms and pave the way for personalized interventions that can delay or prevent neurodegeneration.
Project Reference No. : C5020-25G
Project Title : Sustainable AI Datacenter with Immersion Cooling and Generalized Carbon Estimation
Project Coordinator : Professor Dan WANG
University : The Hong Kong University of Science and Technology
Layman Summary
Recently, an increasing number of AI datacenters have been established to serve AI workloads. AI datacenters use GPUs, and they bring about significant energy consumption and carbon emissions. On the other hand, sustainability has become an important societal issue, and many countries are enforcing carbon taxes.
Past understandings are insufficient for the new generation AI datacenters. In particular, the team saw three new technical opportunities: (1) AI has new computing mechanisms which have yet to sufficiently take energy-efficiency into consideration; (2) there are revolutions in the cooling systems: new immersion cooling systems have much greater heat dissipation efficiency; and (3) time-series foundation models (e.g., weather foundation models) allow general carbon emission estimations even in regions with limited data.
In this project, the team plans to develop (1) new energy-efficient and high-performance computing schemes for AI workloads where the team explores the inefficiencies in the AI computing mechanisms; (2) new energy-efficient computing schemes that maximally explore the merit of immersion cooling systems and avoid their restrictions, (3) new general domain models for carbon forecasting through foundation models and robust carbon-aware computing schemes.
This project is unique since the team not only considers the energy efficiencies of the AI computing mechanisms alone, but also studies the AI computing optimization when taking into consideration the new advances in cooling systems and carbon emission estimations. The team will develop (1) novel cyber-physical co-design schemes to marry the thermal process modeling with the AI computing optimization and (2) novel general domain models to marry the knowledge distilled from weather foundation models with carbon forecasting. Industry is active in this area. The team complements industry research by addressing (1) new inefficiency problems in the AI computing mechanisms with solutions ready to improve their existing approaches, and (2) problems on immersion cooling systems and generic carbon estimations, which will show significance in the long-term.
Project Reference No. : C5085-25G
Project Title : AIoT-powered Multi-modality Underwater SOS System
Project Coordinator : Professor Yuanqing ZHENG
University : The Hong Kong Polytechnic University
Layman Summary
Drowning is the leading cause of accidental death around the world. Many people think that when someone is drowning, it is possible to splash and yell for help, making it easy for lifeguards to notice and provide timely rescue. In reality, however, most drowning victims sink quickly and quietly. Even with lifeguards on duty, some people still drown because help can take too long. The chance of survival increases with earlier rescue efforts.
To assist lifeguards in detecting drowning incidents, some solutions use special cameras deployed above or underwater in swimming pools. These systems trigger an alarm when they detect a swimmer staying still for a long time, but by then, the person might already be in serious trouble. Moreover, camera-based solutions are susceptible to poor lighting conditions, swimmer occlusions, and dynamic backgrounds.
To tackle this urgent issue, this project aims to leverage the recent advances in AI and IoT to develop an AIoT-powered multi-modality underwater SOS system. Instead of solely relying on cameras, the team will design and implement a new underwater SOS communication channel, which allows swimmers to send underwater SOS signals with their waterproof smartwatches in case of drowning or sudden discomfort. In addition, the team will efficiently integrate multi-modality sensor data from smartwatches (e.g., motion and bio-sensors) as well as hydrophones and cameras in swimming pools to provide comprehensive situational awareness for lifeguards and enable accurate and timely drowning event detection.
To this end, the team will develop new solutions to enable reliable underwater acoustic communication with lightweight smartwatches, and accurately locate acoustic transmitters in swimming pools. They will develop a multi-modality sensor fusion framework for efficient and robust drowning event detection. To facilitate multi-modality data collection and performance evaluation, they will instrument swimming pools with sensors and IoT technologies. They will conduct interdisciplinary collaborative research on upgrading the existing swimming pools for swimmer safety, wellness, and sports. It is hoped that the system can be developed, adopted and save people’s lives.
Project Reference No. : C5097-25G
Project Title : Privacy Infrastructure Design for Web3
Project Coordinator : Professor Bin XIAO
University : The Hong Kong Polytechnic University
Layman Summary
Hong Kong is rapidly becoming a global leader in digital finance, driving stablecoin issuance and real-world assets (RWAs) tokenization under its Web3 vision. Yet bringing RWAs onto blockchain depends on secure, accurate physical-to-digital data flows via off-chain oracles and manual checks. Such processes, often based on legacy Web2 systems, rely on centralized data stores and paper proofs, making them vulnerable to manipulation and leaks. In 2024 alone, over 5,000 telecom scams defrauded citizens and academic fraud led to student expulsions from local universities. A digital asset economy must secure data at scale, protecting stakeholders and preserving privacy even against malicious parties in decentralized networks.
To address these challenges, this project pioneers a privacy-preserving Web3 infrastructure that fundamentally redefines data sovereignty in decentralized systems and empowers users to retain full control over their data, while enabling reliable verification without exposing sensitive information. Specifically, the team introduces four breakthroughs to achieve strong privacy without compromising security or compliance: (1) Blockchain infrastructure: The team introduces the first solutions that enable truly anonymous cryptocurrency transfers and cross-chain exchanges, preventing fraudsters from tracking victims’ transactions. (2) Private primitives: The team advances cryptographic primitives tailored for decentralization, facilitating efficient and secure Web3 applications. (3) Private data management: The team proposes the first regulatory-compliant decentralized identity framework that resists fake (Sybil) accounts and ensures compliance with regulations such as anti-money laundering. (4) Web3 applications: The team develops practical solutions to address critical digital trust challenges in Hong Kong, including a collaborative fraud filtering system and a privacy-preserving platform for degree verification.
These technologies ensure that data remains “usable yet invisible”—anti-fraud systems can block scammers without accessing sensitive content, and institutions can verify credentials or authenticity of information without viewing personal information. By replacing opaque central authorities with algorithmic trust, the team returns data ownership to users while still fulfilling regulatory requirements.
This collaborative research aims to advance new theories and techniques for building user-centric (blockchain) systems for secure data sharing and analytics, focusing on decentralized privacy infrastructure for Web3. In partnership with industry collaborators, the team will develop prototypes for anti-scam solutions and degree verification systems. Drawing on the team’s extensive research expertise, strong collaboration record, and dedicated efforts, it is anticipated that this project will deliver impactful results. These outcomes will help rebuild digital trust by making privacy invisible yet unbreakable, demonstrating that fraud can be prevented and regulations can be effectively enforced.
Project Reference No. : C5117-25G
Project Title : Endothelins in mechanoaging and osteoarthritis: biomarker discovery and drug development
Project Coordinator : Professor Chunyi WEN
University : The Hong Kong Polytechnic University
Layman Summary
This project aims to revolutionize the understanding and treatment of knee osteoarthritis (OA), a common and painful joint disease, by focusing on a group of molecules called endothelins. The first goal is to discover if endothelin levels in blood and joint fluid can serve as "biomarkers" in Hong Kong and UK patients, helping doctors predict the severity and progression of OA. Secondly, the team is exploring new therapies by repurposing existing FDA-approved drugs that block endothelin's effects. These drugs will be delivered directly into the knee joint to reduce inflammation and combat joint aging, tested using advanced lab models and animal studies, and potentially combined with other treatments to boost their effectiveness. Finally, the team will investigate a traditional Chinese herbal medicine, Er-Zhi-Wan, to see if its components can also target endothelin pathways for early OA treatment. Ultimately, by tackling OA through biomarker discovery, drug repurposing, and natural remedies, this project seeks to develop more effective, anti-inflammatory, and regenerative treatments for this complex condition.
Project Reference No. : C6004-25G
Project Title : Building Trustworthy Large Language Model (LLM)-Integrated Applications: A Full-Stack Lifecycle Approach
Project Coordinator : Professor Shuai WANG
University : The Hong Kong University of Science and Technology
Layman Summary
Large Language Models (LLMs) and their empowered agents are being rapidly integrated into critical sectors such as finance, education, and public services. However, these applications pose significant risks to security and reliability, including data leaks, biased outputs, and system vulnerabilities. This project aims to build a comprehensive "safety shield" for the entire lifecycle of LLM-integrated applications and agent systems. The team’s "full-stack" approach covers the spectrum from automated testing and supply chain security management during the R&D phase to real-time safety guardrails and trustworthy deployment in the underlying hardware environment during execution. The team will develop vulnerability detection tools, establish "component label" (SBOM) systems to track supply chain risks, and implement real-time protection mechanisms to block unsafe content and dangerous agent behaviors. This research will provide security standards and tools to ensure the reliability of AI applications in critical industrial and public domains.
Project Reference No. : C6024-25G
Project Title : Unraveling Fluid-Solid Interactions and Soft Propeller Mechanics Through Quantitative Optical Imaging
Project Coordinator : Professor Xian CHEN
University : The Hong Kong University of Science and Technology
Layman Summary
This project studies how soft living creatures like jellyfish swim so efficiently in water, and how their soft bodies interact with surrounding fluid. Although scientists understand many parts of fluid mechanics, it is still difficult to directly observe how water flow and soft tissue movement affect each other in real time. To solve this, the team will build an advanced optical imaging platform that can clearly “see” both the moving water and the deformation of soft materials at the same time.
By combining new optical technologies, high-resolution cameras, and innovative light-sheet illumination, the researchers will visualize water flow, measure jellyfish movement, and capture the shape changes of soft tissues. They will first study single jellyfish, then groups, to understand how they coordinate and how vortices and thrust are generated. The team will also create soft artificial propellers to test ideas and validate theories.
The project will deepen scientific understanding of natural swimming, improve experimental tools for studying fluid–solid interactions, and inspire new designs in soft robots, underwater vehicles, medical technologies, and environmental sensing devices.
Project Reference No. : C6041-25G
Project Title : Designing Functionalised Bio-carbons to Propel a Biorenewable Hydrogen Economy
Project Coordinator : Professor Dan TSANG
University : The Hong Kong University of Science and Technology
Layman Summary
Metropolitan cities worldwide face two pressing challenges: increasing food and biomass waste, and a growing demand for clean and renewable energy. In Hong Kong, over 4,000 tonnes of organic waste are thrown away each day, despite its untapped potential as a renewable carbon resource. Meanwhile, the city aims to achieve zero landfill by 2035 and carbon neutrality by 2050. Hydrogen, a clean-burning fuel, is widely regarded as crucial to this transition. However, over 95% of hydrogen globally is still produced from fossil fuels. This gap creates a unique opportunity: Can we turn food and biomass waste into clean hydrogen, sustainably and at scale?
This project aims to transform organic waste into clean fuel and advanced carbon materials. By innovating and integrating biological and thermochemical processes, the team will create a circular and carbon-neutral system that merges cutting-edge science with practical engineering via four Work Packages (WP). WP1 will produce engineered bio-carbons from mixed biomass using microwave-assisted pyrolysis, tuning surface chemistry and porosity for performance. WP2 will apply these bio-carbons to enhance microbes in two-stage anaerobic digestion, boosting hydrogen and methane yields. WP3 will upgrade the biogas by capturing CO2 and converting methane into high-purity hydrogen while co-producing solid carbon, using catalysts derived from the engineered bio-carbons. WP4 will quantify environmental and economic performance and identify deployment and policy pathways for scale-up in Hong Kong.
By integrating biomass waste upcycling, clean energy supply, and advanced material development into one circular system, this approach offers a novel path forward to support multiple policy goals, including waste reduction, hydrogen economy, and climate action plan. This will position Hong Kong as a regional leader in circular economy innovation and provide valuable knowledge for other cities striving to meet climate and sustainability goals.
Project Reference No. : C6053-25G
Project Title : Molecular Mechanisms of Replisome Coupling
Project Coordinator : Professor Yuanliang ZHAI
University : The Hong Kong University of Science and Technology
Layman Summary
In eukaryotes, DNA replication represents a remarkably precise yet extraordinarily complex process that must accomplish both genetic and epigenetic duplication. The replisome—a sophisticated molecular machine—comprises at least three core components: (1) a Cdc45-MCM-GINS (CMG) helicase that unwinds DNA at replication forks, and (2) two distinct DNA polymerases (Pole and Pold) that synthesize complementary strands on the leading and lagging templates, respectively. During S phase, these molecular assemblies processively traverse hundreds of kilobases through chromatin, navigating the topological challenges of nucleosome-packed DNA and other roadblocks. Disruptions in this process can comprise both genetic and epigenetic inheritance, with dire consequences including cancer, aging-related disorders, and developmental abnormalities. This research program will employ an integrated, multidisciplinary approach to elucidate the principles governing replisome coupling. This project will not only advance our fundamental knowledge of genome maintenance but also reveal potential new therapeutic targets for replication-associated diseases.
Project Reference No. : C6078-25G
Project Title : Coordinated Autonomy and Smart Infrastructure for Low Altitude Economy System
Project Coordinator : Professor Fumin ZHANG
University : The Hong Kong University of Science and Technology
Layman Summary
Low-Altitude Economy (LAE) leverages the underutilized airspace below 1000 meters to enable a wide range of applications, including urban goods delivery, emergency response, and infrastructure inspection. As highlighted in Hong Kong’s 2024 Policy Address, LAE represents a transformative opportunity for economic growth, smart city development, and technological leadership. However, the development of LAE faces multiple technical and operational bottlenecks, especially in the high-density urban environments of Hong Kong. Key challenges include ensuring unmanned aerial vehicles (UAVs) safety and reliability in complex cityscapes, managing communications in dynamic and unstable networks, regulating the safety-critical airspace usage, and coordinating large-scale heterogeneous UAV fleets. Existing technologies struggle to address these issues within Hong Kong’s constrained airspace and complex urban environments. This project addresses the core challenges through novel methods and algorithms, including the development of advanced disturbance-rejection control systems for UAVs and a real-time path planning algorithm designed to support parachute-assisted emergency landing. It further proposes a hybrid 5G/LPWAN communication network architecture to ensure resilient connectivity in dynamic environments. The project also introduces a safety-oriented low-altitude airspace structure and configuration design, along with an efficient task allocation framework enhanced by Large Language Models. These interdisciplinary innovations are cohesively integrated within a unified framework and will be rigorously validated in real-world conditions using the HKUST LAE Sandbox, a testbed designed for realistic, policy-driven experimentation and demonstration. A major strength of this project is its strong collaboration with industrial partners, including UAV application operators and communication technology companies. This ensures that the proposed solutions are not only theoretically sound but also practically viable and aligned with industry needs. Furthermore, the project is led by a multidisciplinary team of leading experts in robotics, control theory, network systems, and smart city and urban computing, positioning it for both technical depth and real-world applicability.
Project Reference No. : C6086-25G
Project Title : Federated LPWANs for Scalable IoT
Project Coordinator : Professor Mo LI
University : The Hong Kong University of Science and Technology
Layman Summary
The Internet of Things (IoT) is poised to connect trillions of devices, deepening our understanding of the physical world and underpinning smart city development. Recent advances in Low Power Wide Area Network (LPWAN) technologies enable kilometer-scale coverage and massive device connectivity. However, today’s LPWAN ecosystem remains fragmented: operators compete rather than collaborate, resulting in (1) uncoordinated and congested wireless access that limits effective urban-scale deployments, (2) redundant and costly backhaul investments across operators, and (3) isolated data and resource silos. These challenges significantly constrain the scalability and impact of LPWAN-based IoT systems. This project addresses these gaps by developing key technologies for scalable IoT through a federated LPWAN architecture that enables effective collaboration among heterogeneous networks and operators. The research focuses on improving cross-technology and cross-operator wireless coexistence awareness, and on integrating isolated LPWAN infrastructures via federated backhaul networks. To support experimental validation and real-world adoption, the team will establish a cross-campus federated LPWAN testbed. The testbed will enable interdisciplinary applications such as collaborative building energy and environmental monitoring, and extend IoT connectivity to highly mobile sensing platforms, including UAVs operating in low-altitude airspace.
Project Reference No. : C7026-25G
Project Title : Neural mechanisms of consolidated memory editing in humans and mice
Project Coordinator : Professor S.W.C. LAI
University : The University of Hong Kong
Layman Summary
Maladaptive memories, such as traumatic, fear-related aversive memories, play a central role in the development, persistence, and severity of psychiatric disorders including post-traumatic stress disorder, depression, and anxiety disorders. These maladaptive memories often become involuntary, intrusive, and resistant to natural extinction, maintaining pathological emotional states and impeding recovery. Sleep is critically involved in the consolidation, reconsolidation, and emotional modulation of memories. Targeted memory reactivation (TMR), a non-invasive procedure that delivers sensory cue associated with prior wakeful learning during sleep, has emerged as a promising method to modulate maladaptive memory processing. Project team’s recent findings demonstrate that TMR can exert bi-directional effects on fear memories, either strengthening or weakening memories, depending on the sleep substage in mice. These findings suggest potential strategies to weaken fear memories and promote adaptive memory processing in humans.
This research aims to combine neuroscience, engineering, and artificial intelligence to develop new ways to treat mental health issues. First, the research team will study how reactivating fear memories during sleep affects the brain in mice, using advanced imaging and genetic tools to understand the underlying neural mechanisms. Next, they will test strategies in humans, such as reactivating positive memories to counteract negative ones or directly reducing fear memories during sleep. Finally, they will create a wearable device prototype and a smartphone app that can deliver personalized cues during sleep, making this approach safe and easy to use at home. Taken together, this project seeks to pioneer a new frontier in mental health treatment by leveraging sleep and memory reprocessing to weaken maladaptive memories. The findings could revolutionize psychiatric interventions with non-invasive, personalized, and scalable AI-powered therapies, ultimately advancing our understanding of memory mechanisms and promoting mental well-being.
Project Reference No. : C7042-25G
Project Title : A Deep Dive into Immune Dynamics and Drug Resistance to Advance Therapeutic Innovations for FGFR2-altered intrahepatic cholangiocarcinoma
Project Coordinator : Professor C.C.L. WONG
University : The University of Hong Kong
Layman Summary
Intrahepatic cholangiocarcinoma (iCCA) is a type of liver cancer that arises from the bile ducts within the liver. Many of these tumors have genetic changes in a gene called FGFR2. Although drugs targeting FGFR2 have improved treatment for patients with these genetic alterations, their effectiveness is often limited because the cancer can become resistant, and the tumor environment can suppress the immune system. This project’s goal is to study the molecular, immune, and metabolic effects of FGFR2 changes in iCCA. The team will analyze how the immune system interacts with these tumors and how the cancer evades immune attack. They will also investigate how the cancer adapts its metabolism and develops resistance to FGFR2-targeted drugs. Additionally, they aim to test new immunotherapies that can overcome the immune suppression and develop combination treatments. This research will help create personalized treatment strategies that could significantly improve outcomes for patients with FGFR2-mutated iCCA.
Project Reference No. : C7088-25G
Project Title : Human centromere structure and its maintenance mechanism
Project Coordinator : Professor K. ZHOU
University : The University of Hong Kong
Layman Summary
DNA is the blueprint of life within each individual cell. When cells divide, this blueprint needs to be accurately copied and split between the daughter cells. In human cells, DNA is organized into chromatin, and this project focuses on a critical region of a chromosome called the centromere. It functions like the "handle" to ensure that chromosomes are pulled apart correctly during cell division. Without a properly functioning centromere, cells can receive the wrong amount of DNA, leading to serious problems. In fact, a faulty centromere is a common feature in many cancers and genetic disorders.
In this collaborative proposal, the team aims to understand exactly how this vital "handle" is built, how it stays intact, and how it keeps its identity even when the cell is busy copying DNA or reading its genetic instructions. To achieve this, they are applying various advanced microscopy techniques, including "freezing and slicing" cells through cryo-focus ion beam (cryo-FIB) milling, correlating light and electron microscopy (CLEM), followed by cryo-electron tomography (cryo-ET) to see the centromere's structure up close in its natural state. A particular focus is on the chaperone protein FACT, which plays a crucial role in maintaining the centromere's unique properties. By integrating biochemical analysis and click chemistry, they will examine how FACT interacts with centromeric chromatin and its associated protein networks during different cell cycle phases.
By unravelling these mysteries, this research will provide fundamental insights into how our cells maintain genetic stability throughout the cell cycle. This knowledge could ultimately pave the way for new ways to diagnose and treat diseases like cancer that are caused by errors in cell division.
Project Reference No. : C7112-25G
Project Title : Exploring an adaptive transboundary flood risk governance framework in the Greater Bay Area: An enhanced social-ecological-technological systems approach
Project Coordinator : Professor S. HE
University : The University of Hong Kong
Layman Summary
The Greater Bay Area (GBA) is currently confronting a significant challenge as climate change and rapid urbanization have led to an increase in flood risks that extend beyond city boundaries. Despite the growing urgency, existing flood management systems often remain fragmented and heavily rely on conventional engineering solutions. In response to these challenges, this project proposes an innovative adaptive transboundary governance framework. This framework recognizes that achieving flood resilience requires more than technical solutions; it is the result of the dynamic interaction of social, ecological, and technological factors. By addressing multiple gaps in database, strategies, and institutions between major cities such as Hong Kong, Shenzhen, and Guangzhou, the project aims to foster collaboration and coordination across these jurisdictions, gaining a deeper understanding of why flood vulnerability varies among different population groups, and how to prevent situations where protective measures in one city inadvertently increase risks for neighboring areas.
To bridge theories and practices and to deliver the project’s objectives, the team will develop an advanced data platform that integrates climate and environmental sciences, geographical and urban informatics, and socioeconomic data to precisely project future flood scenarios along with urbanization trends. The project will provide practical solutions, including an interactive website for real-time risk assessment, nature-based urban planning strategies, and enhanced insurance models focused on key development areas such as the Northern Metropolis, Qianhai, and Nansha. The overarching goal is to position the GBA as an international exemplar achieving equitable climate resilience, providing policymakers, businesses, and communities with the necessary knowledge and resources to navigate climate change securely to achieve prosperity and sustainability.
Project Reference No. : C7140-25G
Project Title : The Science of Sustainable Structural Materials: microstructure evolution, simulation tools and experiments
Project Coordinator : Professor D.J. SROLOVITZ
University : The University of Hong Kong
Layman Summary
Modern structural materials, like those used in cars, aircrafts and buildings, are often improved by adding special alloying elements. Many of these elements are not earth-abundant, vulnerable to supply chain risks, expensive, or environmentally harmful to mine and recycle. This project asks: can we instead achieve high performance by optimizing the internal structure of common metals and stabilizing the structure through the use of only very small quantities of alloying elements at locations in the material where they are most effective?
Most engineering metals/alloys contain a high density of “grain boundaries” (i.e., internal interfaces between crystallites/grains). These boundaries, along with other internal structural features such as grain orientation, shape, size and their distribution strongly influence strength, ductility and durability. The central idea of the project is to design and stabilize such microstructures so they deliver outstanding properties, while keeping the overall composition simple and sustainable.
The team will combine advanced computer modelling, machine learning and state of the art experiments to achieve these goals. First, the team will develop predictive simulations that show how internal structure changes can be manipulated through mechanical deformation and heat treating, helping to design processing routes that produce targeted internal structure (for example, layered or gradient structures called heterostructures that blend strength and toughness).
Second, the team will examine which elements can be added, in small amounts, to segregate specifically to grain boundaries to stabilize the architected internal structure. This will be explored mainly in aluminum and titanium based alloys made from earth abundant elements.
Third, the team will develop and validate mechanistic models of how dislocations (the carriers of deformation) interact with grain boundaries, and embed these models in engineering scale simulation tools to predict how complex microstructures behave under in-service conditions.
Finally, the team will synthesize and test model alloys to validate the simulation and model predictions, and use machine learning to iteratively optimize compositions and microstructures. The ultimate aim is the development of a new, science based pathway to design sustainable structural alloys that meet demanding mechanical requirements while reducing cost, reliance on critical elements, and environmental impact.
CRF 2025/26 Collaborative Research Equipment Grant (CREG) Proposals
Project Reference No. : C4018-25E
Project Title : A State-of-art Desktop X-ray Absorption and Emission Testing Platform for Advanced Materials Research in Hong Kong
Project Coordinator : Professor Ye CHEN
University : The Chinese University of Hong Kong
Layman Summary
High-energy X-ray spectroscopy techniques, particularly X-ray absorption spectroscopy (XAS) and X-ray emission spectroscopy (XES), are indispensable tools to characterize materials down to atomic scale for advancing fundamental research in nanoscience, condensed matter, energy storage and conversion, environmental science, and biomedicine. XAS quantifies oxidation states, coordination symmetry, interatomic distances, and neighbor identities, while XES provides ultrahigh-sensitivity analysis of electronic configurations and ligand-specific coordination environments, uniquely capable of distinguishing light-element coordination around transition metals where XAS is ineffective. Historically, these powerful techniques must rely on synchrotron radiation sources, restricting access for regions like Hong Kong due to the prohibitively high infrastructure costs and spatial demands. Consequently, Hong Kong researchers have relied heavily on proposal-based beamtime allocation at overseas synchrotron facilities. Such mode can hardly meet our expanding research needs, primarily due to time-consuming sample testing procedures and inflexible experimental designs.
This proposal establishes a highly integrated benchtop XAS-XES testing platform in Hong Kong. The system features synchrotron-comparable performance (wide energy range, high resolution, low drift, and wide element coverage), in-house operation and customization (reduces analysis time from months to days), and advanced in situ capabilities for real-time, nondestructive monitoring of dynamic processes in solids and liquids. The concurrent XAS-XES measurements deliver cross-validated, atomic-level insights into electronic structures and bonding configurations unattainable with either technique alone. This facility eliminates geographical bottlenecks, empowering Hong Kong’s research community in materials science, green energy, environmental science, and biomedicine. By providing direct access to state-of-the-art characterization, it cements Hong Kong’s position as a leader in next-generation materials and technological innovation.
Project Reference No. : C4032-25E
Project Title : 3D-Scope: A High-precision and Low-latency 3D Indoor Modeling and Localization System for Smart Building, Epidemic Analysis, and Embodied AI
Project Coordinator : Professor Zhenyu YAN
University : The Chinese University of Hong Kong
Layman Summary
Accurate indoor modeling and localization are essential for modern living, as people spend over 85% of their time inside buildings. However, existing technologies struggle to provide the real-time, high-precision 3D indoor information required for advanced healthcare, robotics, and smart building applications due to limitations in latency and accuracy. This project aims to develop "3D-Scope," a novel system capable of creating quick, highly detailed 3D maps and tracking movement with sub-centimeter accuracy. The results of this project will revolutionize the understanding of indoor spaces, offering significant benefits for epidemic modeling, autonomous robotics, and energy-efficient building management in Hong Kong and beyond.
Specifically, the objectives of this project include 1) Development of smart sensing equipment using LiDAR that processes data rapidly for millimeter-level accuracy; 2) A collaborative scanning system that combines fixed sensors with mobile robots to ensure complete room coverage; 3) Creation of "semantic maps" that not only capture geometry but also recognize objects and their relationships, allowing robots to interact intelligently with their surroundings; and 4) Enhancement of navigation technologies to achieve pinpoint location accuracy even in complex or crowded indoor environments; and 5) Validation in diverse indoor environments, where the developed equipment will be extensively tested in real-world scenarios and various indoor venues to verify its applications in epidemic analysis, robotics, and smart building systems.
The proposed 3D-Scope system will provide a new indoor sensing technology by enabling real-time spatial and semantic understanding of environmental details. Its applications are extensive, ranging from improving epidemic modeling and infection control policies to transforming smart building operations and advancing AI-driven robotics. By integrating cutting-edge technologies such as LiDAR sensing, edge computing, and artificial intelligence, this project will create a versatile platform that generates significant social and economic impact for Hong Kong and the broader region.
Project Reference No. : C4056-25E
Project Title : The First State-of-the-Art Molecular Resolution Light Microscope for Live Cell Imaging into our Advanced Imaging Platforms to Timely Promote Interdisciplinary and Advanced Life Sciences Research in Hong Kong and Beyond
Project Coordinator : Professor Liwen JIANG
University : The Chinese University of Hong Kong
Layman Summary
The newly developed MINFLUX (Minimal Photon Fluxes) microscope achieves live-cell imaging of molecular resolution down to 1-3 nm by localizing single fluorophores via rapid probing with a donut-shaped beam that features a local intensity minimum. This is currently the highest achievable resolution for live cell imaging using light microscope. MINFLUX has become a very powerful and essential tool for investigating cellular processes and mechanisms at molecular level. Here the team proposes to timely establish the first state-of-the-art molecular resolution light microscope for live cell imaging in Hong Kong by acquisition of a MINFLUX microscope and related Software/algorithms for image analysis and application. The timely establishment, integration, further development and application of MINFLUX in this interdisciplinary research will be a great supplement to the existing imaging platforms and make the Advanced Life Sciences Research reach the next highest levels in Hong Kong and beyond.
Project Reference No. : C5081-25E
Project Title : Advanced Single-Particle Mass Spectrometer to Uncover Hidden Aerosol Complexity Impacting Health and Climate
Project Coordinator : Professor Ling JIN
University : The Hong Kong Polytechnic University
Layman Summary
Air pollution remains a pressing global challenge, contributing to millions of premature deaths each year and exacerbating climate instability. In Hong Kong and the Greater Bay Area (GBA), complex pollution sources including maritime emissions, urban traffic exhaust, and transboundary wildfire smoke pose substantial risks to public health and the environment. Despite significant advances in air quality research, conventional bulk analysis methods oversimplify aerosol systems by averaging chemical compositions across entire particle populations. This approach neglects the aerosol mixing state, defined as the heterogeneous distribution of chemical components within individual particles, which critically governs particle toxicity, climate interactions, and the effectiveness of mitigation strategies. Addressing this gap requires advanced single particle analysis capable of resolving aerosol complexity.
This proposal seeks funding to acquire an advanced Single Particle Mass Spectrometer (SPMS) equipped with dual ionization technologies, namely laser desorption and ionization and resonance enhanced multiphoton ionization. These state-of-the-art capabilities will enable real-time, high-resolution detection of both refractory and non-refractory aerosol components, including metals, soot, secondary organics, and polycyclic aromatic hydrocarbons, which are not fully accessible with conventional online instruments. By resolving aerosol mixing states, the proposed SPMS will provide transformative insights into aerosol health effects, climate impacts, and pollution sources.
The instrument will catalyse advances in three key research areas. First, it will improve understanding of aerosol toxicity by linking chemical heterogeneity to differential health outcomes. Second, it will refine climate models through detailed characterisation of aerosol radiative and cloud forming properties. Third, it will enhance pollution source apportionment by leveraging novel tracers such as metal-organic complexes and polycyclic aromatic hydrocarbon signatures to disentangle overlapping emission profiles. The instrument will be deployed at roadside stations, urban supersites, and regional background sites in collaboration with the Hong Kong Environmental Protection Department. It will also support broader initiatives across the GBA and international field campaigns through partnerships with global collaborators.
Complementing the research programme, a dedicated education and training plan will support graduate students, postdoctoral researchers, and early career faculty through the SPMS Eurasia Course and Annual Symposium. These activities will provide hands on training in instrument operation, data analysis, and policy-relevant applications, fostering a skilled and interdisciplinary workforce for air quality research in Hong Kong and beyond. By overcoming the limitations of bulk aerosol analysis and focusing on particle mixing states, this initiative will deliver transformative outcomes, including improved health risk assessments, enhanced climate modelling, and evidence-based air quality management strategies. In doing so, Hong Kong will further consolidate its leadership in atmospheric science while addressing critical regional and global challenges in air quality and climate mitigation.
Project Reference No. : C5082-25E
Project Title : A trimodal PET/SPECT/CT animal imaging system for molecular imaging and radiopharmaceutical research
Project Coordinator : Professor Jung Sun YOO
University : The Hong Kong Polytechnic University
Layman Summary
This proposal aims to establish Hong Kong’s first nuclear molecular imaging facility by acquiring a trimodal Positron Emission Tomography (PET)/Single-Photon Emission Computed Tomography (SPECT)/Computed Tomography (CT) animal imager, addressing a critical gap in the region’s biomedical research infrastructure. Integrating PET, SPECT, and CT in a single platform provides unparalleled sensitivity, spatial resolution, multi-energy detection, and anatomical correlation, enabling real-time, quantitative assessment of drug biodistribution in animal models.
The facility will serve as a collaborative hub for multidisciplinary teams from academia, hospitals, and industry, fostering impactful research and training the next generation of experts in radiochemistry, nuclear imaging, and related fields. Key research applications include the development and in vivo validation of novel diagnostic and therapeutic radiopharmaceuticals, noninvasive longitudinal studies of disease progression and therapeutic response, and the integration of artificial intelligence for advanced image analysis.
By establishing this advanced imaging platform, Hong Kong SAR will be positioned at the forefront of molecular imaging and radiopharmaceutical research in the region. Ultimately, this initiative will contribute to the advancement of precision medicine, improvement of patient care, and the strengthening of Hong Kong’s role as a leader in biomedical innovation.
Project Reference No. : C6054-25E
Project Title : Development of an Automated UAV Vertiport for Low Altitude Economy Research at HKUST
Project Coordinator : Professor Zili MENG
University : The Hong Kong University of Science and Technology
Layman Summary
The Low Altitude Economy (LAE) is rapidly emerging as a significant sector nowadays. Recognizing its potential, the Government of the Hong Kong Special Administrative Region, as highlighted in the 2024 Policy Address, is actively formulating strategies to leverage LAE for economic growth. Key considerations involve air traffic management, communication for dense UAV operations, safety, and regulatory compliance, which necessitate comprehensive research. The Hong Kong University of Science and Technology (HKUST) has taken a proactive role in advancing LAE with the establishment of the Low Altitude Economy Research Center (LAERC) in January 2025. The center aims to bridge academic research, industry applications, and governmental policies, fostering an ecosystem for developing cutting-edge technologies and translating research into practical solutions. Notably, HKUST has gained approval for a regulatory sandbox, allowing for expansive UAV testing using the university's campus and surrounding areas.
Central to advancing LAE research is the proposal for an automated UAV vertiport, a critical infrastructure envisioned to support and accelerate R&D by minimizing human error, enhancing operational efficiency, and providing a scalable testing environment. This automated system would reduce risks associated with manual operations, enable continuous operations, and facilitate large-scale experiments crucial for studying UAV coordination, scheduling, and data integration.
Despite Hong Kong's technological landscape, there is currently no automated UAV vertiport dedicated to advanced research, with existing infrastructure being largely ad-hoc. The proposed facility at HKUST aims to address this gap, assisting in empirical research into operational safety, efficiency, scalability, and airspace integration challenges within Hong Kong's complex environment. This initiative, therefore, holds the promise of driving innovation in LAE technologies and contributing significantly to the sustained growth of the sector in Hong Kong.
Project Reference No. : C6076-25E
Project Title : A Biaxial Centrifuge Shaker for Enhancing Research on Seismic and Climate Resilience of Infrastructure
Project Coordinator : Professor Gang WANG
University : The Hong Kong University of Science and Technology
Layman Summary
Centrifuge modeling is a powerful and advanced laboratory technique for geotechnical engineering. By using centrifugal force, a centrifuge creates a stress field equivalent to that of a full-scale (prototype) structure within a smaller, reduced-scale model, making it a cost-effective and time-efficient approach to studying complex geotechnical phenomena. Established in 2001, the Geotechnical Centrifuge Facility (GCF) at HKUST has become the world's leading laboratory in the field of geotechnical engineering. Presently, GCF houses a 400 g-ton beam centrifuge and an 850 g-ton drum centrifuge, which are among the most powerful facilities in the world and are unique in Hong Kong SAR. The beam centrifuge is also equipped with the world's first biaxial shaker for dynamic testing. However, after 25-years of operation, the existing shaker has aged, and cannot be fully functional without extensive repair. Moreover, the shaker can only accommodate a small model size and payload. Such limited capacity is insufficient for modeling a spectrum of dynamic problems for larger infrastructure, such as seismic soil-structure interaction, design of high slopes, deep tunnels and large dams etc. against earthquake load.
Due to significant limitations of the existing shaker, researchers from five universities in Hong Kong and abroad (HKUST, HKU, PolyU, Shenzhen U, UC Berkeley) jointly propose to develop a new, upgraded biaxial shaker with a larger table size. The team also propose to construct an environmental simulator to integrate climate considerations (controlled temperature, humidity, rainfalls etc.) into centrifuge shaking table tests, making it a unique facility to study seismic and climate resilience of key infrastructure.
The shaker will be installed onto the existing beam centrifuge, operated and maintained by the experienced engineers and technicians at GCF. Together with existing facilities, the new biaxial shaker could significantly improve research capability on seismic safety evaluation and mitigation of slopes, buildings, tunnels, offshore structures and other key infrastructure, with the consideration of climate change. It can also help Hong Kong to develop its own seismic code, and provides a platform for local and international researchers/engineers to address a variety of challenges related to seismic and climate resilience design of infrastructure, and thus enhance Hong Kong's role as one of the world-leading centrifuge research centers.
Project Reference No. : C7017-25E
Project Title : An angle-resolved ultrafast cathodoluminescent microscope for interdisciplinary studies of quantum optics and quantum materials
Project Coordinator : Professor Y. YANG
University : The University of Hong Kong
Layman Summary
This project will develop an electron microscope to study how tiny beams of electrons interact with light and advanced materials. Instead of focusing on very high energies, the team uses gentle, low energy electrons and ultrafast laser pulses. This could make it easier to “see” subtle quantum effects that are usually hidden in conventional instruments.
By combining electron imaging, light emission measurements, and precise timing in a single setup, the system can record where, when, and how energy moves between electrons, light, and materials at the nanometre scale. This opens the door to new ways of designing quantum devices, compact radiation sources, and energy-efficient electronic and photonic components.
The project brings together expertise in electron microscopy, lasers, and quantum materials, and will help position Hong Kong as a regional centre for next generation quantum instrumentation and technology.
CRF 2025/26 Young Collaborative Research Grant (YCRG) Proposals
Project Reference No. : C1068-25Y
Project Title : Electrosynthesis of high-value organonitrogen compounds via C-N coupling: from catalyst design to cell engineering
Project Coordinator : Professor Zhanxi FAN
University : City University of Hong Kong
Layman Summary
Electrosynthesis using renewable electricity offers a transformative pathway to produce high-value organonitrogen compounds from abundant small molecules like CO2 and NO3-, fundamentally changing how we approach chemical manufacturing. Traditional production of these essential chemicals, such as urea (>180 million tons annually) and cyclohexanone oxime (key intermediate for nylon-6), relies on energy-intensive thermochemical routes involving fossil fuel-derived feedstocks under harsh conditions. In stark contrast, electrochemical approaches enable ambient-condition operations with precise potential control, eliminate the need for high-pressure equipment, reduce safety hazards, and offer modular scalability for distributed manufacturing. These processes can achieve near-zero carbon emissions when powered by renewable energy sources. However, current electrochemical coupling reactions face significant challenges including insufficient activity, poor selectivity, low Faradaic efficiency, and catalyst instability, primarily due to complex multi-electron reaction mechanisms and inadequate understanding of structure-performance relationships. To address these critical limitations, this proposal presents a comprehensive strategy by integrating catalyst design, mechanism study and cell engineering, aiming to provide crucial technological foundations for sustainable chemical manufacturing.
Project Reference No. : C1085-25Y
Project Title : Nearshore intensification and inland impact of the tropical cyclones in the Western North Pacific under global warming
Project Coordinator : Professor Jung-Eun CHU
University : City University of Hong Kong
Layman Summary
Tropical cyclones (TCs) are powerful storms that cause major damage worldwide. Global warming is making their impacts worse, especially in coastal areas like southern and eastern China. Recent studies show that these storms are now intensifying more quickly near shore, gaining strength just before landfall. This means they stay stronger for longer after moving inland, leading to heavier rainfall, flooding, and storm surges. Events such as Typhoon Saola and Typhoon Koinu in 2023 highlight the urgent need to understand why this nearshore intensification is happening.
This project focuses on three goals: uncovering the physical reasons behind this nearshore intensification, testing whether advanced climate models can accurately represent these processes, and assessing risks to Hong Kong’s infrastructure. By combining climate science, engineering, and high-resolution modeling, the team aims to provide practical insights for building safer cities. The methods developed here can also help other coastal megacities prepare for future storm risks.
Project Reference No. : C4003-25Y
Project Title : Cross-Cultural Analytics: Using Big Data, Large Language Models, Knowledge Graphs and Rare Books to Chart the Influence of Chinese Culture in the West, 1550-1900
Project Coordinator : Professor Stuart Michael MCMANUS
University : The Chinese University of Hong Kong
Layman Summary
This project aims to make Hong Kong a global leader in a new field called Cross-Cultural Analytics, which uses big data and AI to study how cultures have interacted over time. The team will focus on how Chinese ideas influenced Europe between 1550 and 1900—a period of major change. Instead of looking at isolated cases, the team will use advanced computational tools and techniques to analyse thousands of historical texts, including rare books from CUHK Special Collections and massive online databases. The team combines expertise in computer science, linguistics, and history to trace how Chinese philosophy, science, medicine, literature, and historical writing shaped Western thought. Using cutting-edge techniques, the team will uncover patterns of cultural exchange at scale. The project will produce research papers across a range of disciplines, a public exhibition, and lay the foundation for a permanent research centre in Hong Kong, strengthening its role as a hub for East-West collaboration.
Project Reference No. : C4069-25Y
Project Title : Accelerating Silicon/Perovskite Tandem Photovoltaics Readiness through Phase-Stable and Hot-Spot Resistant Perovskite Design (STAR-PV)
Project Coordinator : Professor Martin STOLTERFOHT
University : The Chinese University of Hong Kong
Layman Summary
The solar cell industry is rapidly evolving to address global energy demands, with Si/perovskite tandem solar cells emerging as a leading technology due to their ability to surpass the efficiency limits of traditional silicon cells (34.9% vs. 27.8%). However, to achieve widespread commercialization, the Si/perovskite tandem modules must achieve a similar stability to that of traditional Si modules (>25 years) while delivering significantly higher performances at minimal extra costs. Although perovskites have made remarkable improvements in terms of operational lifetimes, critical stability challenges, particularly for wide-bandgap (WBG) perovskite top cells, remain. This includes (1) phase instability and degradation due to ion migration in WBG perovskites, (2) mechanical instability, e.g., due to delamination, and (3) reverse-bias breakdown and hotspot formation. STAR-PV brings together a young, energetic team from 3 HK and 1 mainland University, as well as leading industry partners such as LONGi and Trina Solar, to overcome these challenges through fundamental research in a collaborative manner. By systematically resolving these stability bottlenecks, the team aims to achieve record-setting tandem cells with >34% efficiency and projected operational lifetimes of 25 years, as well as all-around stability under photothermal, thermal cycling, and reverse bias stressors. The results will pave the way for scalable, industry-compatible manufacturing.
Project Reference No. : C4109-25Y
Project Title : Collaborative Intelligence and Autonomy through Cost-effective Embodied AI in Robotic Surgery
Project Coordinator : Professor Qi DOU
University : The Chinese University of Hong Kong
Layman Summary
Surgical robot automation is developing rapidly. Just as assistive-driving became standard in new cars and humanoid robots prepare to enter homes, smart surgical robots will eventually reach a stage when AI-assistive autonomous capabilities will enter every operating room, facilitating every surgery. Given the recent success of AI such as foundation models and embodied AI, this moment may come sooner than humans imagine. To be ready for such revolution towards widespread applications of AI-assisted surgical robot in clinical practice, we should significantly reduce the existing developmental cost and address key challenges: 1) Data cost: Current surgical simulation approaches are prohibitively tedious and time-consuming, making it expensive to create diverse training scenarios needed for embodied AI. 2) Skill learning scale-up cost: Surgical skills are complex and require expert demonstrations, making it hard to learn versatile surgical skills for general-purpose task automation. 3) Human-robot collaboration and clinical deployment cost: There is yet limited exploration about how to effectively implement human-robot collaboration in live surgical settings, leaving the deployment cost poorly understood.
In this project, the team aims to innovate low-cost, highly-efficient, human-collaborative solutions that revolutionize AI-assisted automation for surgical robot. To achieve this goal, the team will develop interconnected novel techniques for cost-effective data generation, control policy learning and human robot collaboration. First, the team will develop a fully data-driven surgical environment simulation system leveraging 3D Gaussian scene representation and physics-based soft tissue interaction. It will significantly reduce the data generation cost while improving simulation diversity and realism. Second, taking advantage of the simulator, the team will develop a breakthrough cost-effective and highly-accurate framework of vision-based robot learning which enables zero-shot sim-to-real transfer. The team’s preliminary work has successfully automated various surgical tasks including tissue retraction and vessel clipping under in vivo settings. Third, the team will study trusted human-robot collaborative control mechanisms via efficient LLM-based voice interaction, context-aware intention prediction, and personalized user adaptation, allowing surgeons to naturally collaborate with the AI-automated robot. Lastly, the team will integrate the AI algorithms to a clinically-viable surgical robot platform and validate it with comprehensive experiments on live animal trials using facilities at Multi-scale Medical Robotics Center at CUHK.
Overall, this is a multidisciplinary project with a team of experts in AI, robotics, and surgery. The outcome of this project will transform surgical automation by making it more accessible and human-collaborative, accelerating widespread adoption of AI assistance in robotic surgery, and ultimately reducing surgical waiting time and medical cost for patients.
Project Reference No. : C5101-25Y
Project Title : Smart Environment Engineering via Beyond Diagonal Reconfigurable Intelligent Surface for Future Wireless Networks
Project Coordinator : Professor Shuowen ZHANG
University : The Hong Kong Polytechnic University
Layman Summary
Future wireless networks will support unprecedentedly high data rate and new sensing functions. A key obstacle in realizing this goal lies in the randomness of the wireless propagation environment with possible blockage and severe fading. To address this issue, this project pursues a disruptive approach of smartly engineering (tuning) the wireless propagation environment for meeting the high demands of future networks, via an innovative technology termed beyond diagonal reconfigurable intelligent surface (BD-RIS).
BD-RIS consists of planar meta-surface(s) mounted with massive elements, which can collaboratively introduce controllable changes (or “scattering”) to the incident signals. By properly designing the scattering matrix, the environment can be smartly engineered (tuned) to favour wireless functions. Compared with conventional (diagonal) RIS with a diagonal scattering matrix since each element is separately controlled, BD-RIS scattering matrices can be designed beyond the diagonal structure under less-stringent constraints due to the existence of inter-element connections, thereby offering significantly enhanced design flexibility. This motivates this project to conduct a comprehensive study to unveil and realize the full potential of smart environment engineering via deploying BD-RISs in future networks.
The team gathers experts from both theoretical and experimental fields of wireless technology. By synthesizing the team’s complementary expertise, this project will pave the way for large-scale deployment of BD-RISs in future wireless networks, which will provide enhanced wireless service at low cost, and consequently narrow the digital divide.
Project Reference No. : C5132-25Y
Project Title : Digital Solutions to Manage the Risks of Electric Vehicle Batteries and A Deep Learning Based Index Insurance Contract Design
Project Coordinator : Professor Qin WANG
University : The Hong Kong Polytechnic University
Layman Summary
Electric vehicles (EVs) play a pivotal role in advancing Hong Kong’s carbon neutrality goals. In 2024, over 70% of newly sold vehicles in the city were EVs. However, this swift development brings two key challenges. First, EVs pose distinct fire risks, mainly due to the potential for lithium-ion batteries to overheat or catch fire if damaged or improperly managed. There is a pressing need for more accurate methods to assess battery health and manage the risks of EVs. Existing laboratory-based diagnostic approaches often do not reflect the actual conditions experienced by EV batteries in real-world use, limiting their effectiveness. Second, both insurance companies and EV dealers currently lack effective actuarial tools to accurately assess and set appropriate insurance policies for EVs. There is a strong need for effective insurance pricing strategies that adequately compensate EV drivers for battery-related losses. Addressing these issues is essential to support the continued safe and sustainable expansion of electric mobility in Hong Kong.
The primary objectives of this research are to develop innovative digital tools for assessing EV battery health without the need for disassembly, and to design insurance contracts for insurers and EV dealers that account for both physical and financial considerations. First, the team will generate multi-modal battery excitation profiles to facilitate the testing of a commercial EV battery pack in the laboratory. In order to encompasses the full operational and failure envelope of modern battery systems, the team will integrate the multi-modal excitation profiles with meticulously controlled extreme condition testing. Second, the team will develop a transferable mutual information-based diagnostic methodology for EV batteries fault detection, as well as an advanced multi-task dual-transformer framework for remaining useful life (RUL) prediction of EV batteries in small sample condition. Third, the team will develop a risk evaluation model that quantifies EV battery risks and helps EV fleet operators optimize battery replacement and EV operations. Fourth, the team will adopt a deep learning-based approach to design index insurance for EV batteries, and develop a simplified model to comply with regulatory requirements.
The final deliverables of this project include a high-fidelity EV battery testing lab, an open-source EV health diagnostic dataset, an innovative digital platform for real-world EV battery fault detection and RUL prediction, a risk management tool, an index insurance product, and recommendations to the government regarding the regulations of EV risk management and insurance policy.
Project Reference No. : C6002-25Y
Project Title : Extreme-mission Microelectronic Passive Transducers
Project Coordinator : Professor Yansong YANG
University : The Hong Kong University of Science and Technology
Layman Summary
Microelectronic transducers play a critical role in extreme-environment applications, such as deep space exploration, deep Earth sensing, and cryogenic quantum technologies, where conventional semiconductor-based devices fail due to thermal, mechanical, and radiative stress. Semiconductor transducers suffer from intrinsic carrier freeze-out at cryogenic temperatures, excessive leakage at high temperatures, and radiation-induced degradation, thereby limiting their operational range and long-term reliability. To overcome these limitations, this project proposes a new class of passive piezoelectric microelectronic transducers based on ultrawide-bandgap (UWBG) materials that function without external power or active electronic circuitry.
The project centers on scandium-doped (alloyed) aluminum nitride (Al1-xScxN or AlScN), a UWBG piezoelectric material with a high Curie temperature, radiation hardness, and tunable electromechanical properties. Although the relationship between Sc concentration and piezoelectric enhancement has been established, its thermal and radiative stability across extreme environments remains poorly understood. This study aims to investigate the temperature- and radiation-induced evolution of polarization, charge transport, and electromechanical coupling in AlScN, with the goal of identifying the intrinsic limits to its long-term functionality.
To achieve device-level implementation, the project introduces a novel fixed-fixed acoustic boundary heterostructure, integrating AlScN with silicon carbide (SiC) and low-temperature chemical vapor deposition (CVD) diamonds. This multilayer stack leverages the thermal conductivity, acoustic confinement, and mechanical stability of SiC and diamond while addressing practical packaging limitations for extreme missions. The heterointerfaces, however, present new challenges such as thermal expansion mismatch, interfacial stress, and energy dissipation. Low-temperature diamond deposition techniques (<700°C) will be developed to mitigate the abovementioned challenges, thereby significantly enhancing transducer reliability.
Key failure mechanisms, including thermal-assisted acoustomigration, will be studied through combined experiments and simulations to reveal atomic-scale degradation pathways under high temperature and high acoustic stress. Systematic lifetime testing under thermal cycling, mechanical stress, and gamma irradiation will be performed to identify critical thresholds for performance degradation.
Through fundamental investigations of material behavior, interfacial physics, and heterostructure, this project aims to establish a physical framework for high-temperature, radiation-resistant piezoelectric transduction. The results will provide design guidelines for UWBG transducers and contribute to the broader understanding of functional materials under extreme conditions. By bridging materials physics and transducer design, the project seeks to lay the scientific foundation for the next generation of resilient, high-frequency microelectronic systems for planetary exploration, subsurface monitoring, and beyond.
Project Reference No. : C6011-25Y
Project Title : Control of Chirality and Spin in Halide Perovskites: from Fundamental Study to Quantum Sensing and Optoelectronics
Project Coordinator : Professor Haipeng LU
University : The Hong Kong University of Science and Technology
Layman Summary
The semiconductor industry is essential for modern optoelectronics but faces challenges in speed and energy efficiency. To improve these aspects, researchers are exploring new methods, particularly spin-based technologies. This proposal focuses on the development of chiral metal-halide semiconductors (MHS) which utilize a phenomenon called chiral-induced spin selectivity (CISS) to better manipulate spins at room temperature. Prior studies have shown that CISS can result in "spin-polarized" excitons and electrons, particularly in chiral perovskites. However, there are still significant challenges to overcome, such as: limited chirality transfer strategies, subpar chiroptical responses, low spin polarization, and an incomplete understanding of how spin polarization works.
To tackle these issues, the project will adopt a multidisciplinary approach, involving: (1) Developing new methods for creating chiral halide perovskites with better spin properties. (2) Using advanced techniques to study the relationship between structure and spin. (3) Investigating the CISS mechanism through chiral materials and quantum sensing. (4) Making efficient spin-optoelectronic devices.
The goal is to unlock the potential of chiral MHS for future optoelectronic and quantum technologies, leading to faster devices that consume less energy. By gaining a deeper understanding of chiral spin dynamics, the project aims to significantly advance the fields of spintronics and quantum information science.
Project Reference No. : C6088-25Y
Project Title : DeepSparseCT: An Ultra-Sparse X-Ray-Driven Foundation Model for Intraoperative 3D CT Reconstruction
Project Coordinator : Professor Xiaomeng LI
University : The Hong Kong University of Science and Technology
Layman Summary
In 2022, there were more than 9 million orthopedic procedures performed in Chinese Mainland and Hong Kong. This number has been steadily increasing at a rapid rate of 15% per year due to factors like an aging population and improved access to healthcare. Currently, computed tomography (CT) scans are used in the clinical workflow to visualize the internal 3D anatomy of the body. CT scans capture a series of X-rays from various angles, resulting in higher radiation exposure and acquisition costs compared to 2D X-rays. During intraoperative procedures, the high radiation dose limits the application of CT scans, forcing surgeons to rely on 2D X-rays for real-time visualization, guidance, and error prevention. However, the two-dimensional nature of X-rays makes it challenging for surgeons to accurately assess the three-dimensional shape of complex anatomy during surgery. This can lead to a misleading perception of orthopedic implant positioning, potentially resulting in surgical failure. Therefore, developing an intraoperative 3D CT imaging system derived from ultra-sparse X-rays - featuring reduced radiation doses and the capability for precise implant placement - holds great promise for significantly improving treatment outcomes.
Developing methods for ultra-sparse-view-X-ray-driven reconstruction remains an open challenge. This project will develop DeepSparseCT, a foundational AI platform enabling surgical-grade ultra-sparse CBCT reconstruction. It serves as a revolutionary low-radiation-dose and intraoperative 3D imaging visualization platform for orthopedic surgery via ultra-sparse-view X-rays. This can be achieved by pioneering novel approaches for volumetric CT reconstruction. Firstly, the team will collect a large-scale cross-regional CT scans dataset that covers various anatomy regions. Secondly, the team will propose the first cross-regional foundation model for ultra-sparse-view CBCT reconstruction. Thirdly, the team will propose the surgical guidance-oriented regional semantic fine-tuning, which innovatively integrates surgeon-annotated visual-language supervision with intraoperative decision-making logic, to suppress spurious artifacts while amplifying critical features for intraoperative navigation. Finally, the team will conduct a comprehensive surgical practice evaluation, to verify the clinical utility of the proposed AI platform in terms of real-time intraoperative reconstruction capabilities.
Through achieving these objectives, the project aims to deliver a patient-friendly intraoperative 3D CT imaging system that revolutionizes orthopaedic surgery by leveraging the complementary strengths of the research team.
Project Reference No. : C7058-25Y
Project Title : Molecular Mechanisms of Non-Homologous End Joining Regulating Nucleosome Remodeling in Human DNA Repair
Project Coordinator : Professor S. LIANG
University : The University of Hong Kong
Layman Summary
DNA is the main carrier of genetic information in human cells. While maintaining DNA integrity is essential, DNA damage is inevitable due to various external and internal factors (e.g., radiation, environmental pollutants, DNA replication stress etc.). Among different types of DNA damage, DNA double-strand break (DSB) is arguably the most toxic and deleterious one. In mammalian cells, one major pathway for repairing DSBs is non-homologous end joining (NHEJ), accounting for approximately 75–90% of cases. Given its central role in the DNA damage response and repair (DDR&R) system and in preserving genome stability, understanding how NHEJ is regulated within the context of chromatin and nucleosomes is a fundamental biological question. This collaborative research project will comprehensively investigate the biochemical, biophysical, structural, and functional mechanisms by which human NHEJ regulates nucleosome remodeling during DNA repair. This project will reveal novel molecular insights into NHEJ function on chromatin, providing a deeper fundamental understanding of DDR&R, the maintenance of genome stability, and the regulation of chromatin reorganization.
Project Reference No. : C7065-25Y
Project Title : Investigation the Mechanism of Coronavirus RNA Replication and Translocation within Replication Organelle
Project Coordinator : Professor T. NI
University : The University of Hong Kong
Layman Summary
Coronaviruses hijack host membranes to form "double-membrane vesicles" (DMVs), acting as viral factories where RNA is copied and immunity is evaded. A critical component of these DMVs are pore complexes, which regulate the import of cellular metabolites and the export of newly synthesized viral RNA. While the team’s previous studies revealed the SARS-CoV-2 pore structure, the precise mechanism of RNA translocation and its coordination with the viral RNA-replication machinery remains a key mystery. This project employs a multidisciplinary approach combining molecular virology, in situ structural biology, and advanced microscopy. This project’s primary goals are to resolve the first high-resolution structure of the complete "Replicase-DMV pore complex" and elucidate how viral RNA is threaded through the pore during synthesis. The team will also perform comparative structural analyses of DMV pores from diverse coronaviruses to understand evolutionary adaptations impacting pore function and RNA transport efficiency. This research will profoundly advance our mechanistic understanding of coronavirus replication, focusing on RNA translocation—a conserved and vulnerable process across many RNA viruses. By revealing these fundamental steps, this project’s findings will directly inform the rational design of broad-spectrum antiviral therapeutics. These insights are crucial for developing drugs that disrupt viral factory formation, block RNA synthesis, or impair RNA export, offering a powerful strategy against current and future RNA virus threats.
Project Reference No. : C7143-25Y
Project Title : Wafer-scale integrated III-V/lithium niobate lasers and photonic circuits
Project Coordinator : Professor C. XIANG
University : The University of Hong Kong
Layman Summary
Integrated photonics is revolutionizing technology by using light instead of electricity to process information on small chips. While materials like thin-film lithium niobate (TFLN) are excellent for moving light signals at high speeds, they cannot generate light themselves. Consequently, current systems rely on bulky external lasers, which limits their size and efficiency. This project aims to solve this limitation by developing a new manufacturing platform that combines the best properties of different materials into a single "super-chip".
The research team will use a unique "heterogeneous integration" technique to bond light-generating semiconductor materials onto high-speed lithium niobate, using a silicon layer as a bridge. This approach allows for the creation of powerful, compact devices that were previously impossible to manufacture at scale. Specifically, the project will build ultrafast tunable lasers that can change color rapidly and "frequency combs" that act as precise rulers for light. These advancements will lead to smaller, energy-efficient optical chips, paving the way for faster internet communications, advanced data centers, and high-precision medical imaging systems.