Lingnan University wins 14 awards at 51st International Exhibition of Inventions Geneva - More than twice as many as last year

Awards

Gold Medal with Congratulations of the Jury;

"Thailand Award for the Best International Invention and Innovation" issued by the National Research Council of Thailand

Project Title

FlexiBot: Make Every Parking Spot a Charging Spot

Winning Lingnan Faculty and Research Staff

• Prof Tang Xiaopeng, Assistant Professor (Presidential Early Career Scholar) of the Division of Science
• Mr Wang Fangyu, Senior Research Assistant of the Division of Science

Project description

FlexiBot is an intelligent robot designed to provide wireless charging for electric vehicles in any parking space. A smart solution increasing the efficient use of fixed charging stations, the robot automatically locates and moves to the nearest charging point. Its wireless charging module and adapter are compatible with all electric vehicles, offering park-and-charge compatible chargers. The system also features smart scheduling for charging, automatically avoiding peak electricity times to reduce grid load and cost.

Key Features:

• Autonomous Mobile Charging

The robot goes to the vehicles to be charged automatically, offering flexibility and reducing the cost of parking modifications.

• Excellent Compatibility

Uses an adapter suitable for all electric vehicles, eliminating compatibility issues and ensuring safe and convenient operation.

• Smart Scheduling

Automatically arranges charging times and plans charging queues to optimise energy usage.

Awards

Gold

Project Title

SmartCool: Digital Twin AI, Cool Smarter with Human Feedback, Hit Carbon Zero

Winning Lingnan Faculty and Research Staff

• Prof S. Joe Qin, President and Wai Kee Kau Chair Professor of Data Science
• Prof Mo Yanfang, Assistant Professor of the Division of Industrial Data Science of the School of Data Science
• Dr Liu Yiren, Postdoctoral Fellow of the Division of Industrial Data Science of the School of Data Science
• Mr Li Jicheng, PhD student at the School of Data Science
• Mr Zhu Zhongxi, PhD student at the School of Data Science

Project description

The SmartCool project is an AI-driven air conditioning management system that optimises performance through the integration of DEMMFL predictive algorithms and RL-MPC control technology. The system simulates both the supply side (chiller units) and the demand side (office spaces) at the same time. Smart feedback devices on office desks instantly report their thermal sensation, enabling precise optimisation of the office air conditioning. This ensures personalised comfort as well as systematic, efficient energy savings. Through accurate load forecasting and reinforcement learning scheduling, SmartCool achieves a precise match between cooling supply and demand. This not only reduces energy consumption, but also resolves the familiar issue of office overcooling, creating a human-centric intelligent environment.

Key Features:

• Desk-side Smart Feedback Device

A personalised tool (M5DIAL) on each desk encourages users to provide real-time feedback on their thermal comfort. This input is used to adjust the air conditioning settings in their immediate vicinity for accurate and individualised comfort.

• Simplified Regulation

The system integrates all user feedback through a reinforcement learning (RL) algorithm to dynamically optimise the settings of each air conditioning terminal unit, which enhances overall comfort while resolving the usual office overcooling pain point, optimising energy savings and personal comfort.

• Forecasting for Energy-Saving

The system accurately determines the actual cooling load required, and by integrating the DEMMFL forecasting method, it achieves long-term, hour-by-hour load predictions based on the weather, ensuring that the energy supply exactly matches demand, and reducing waste at source.

Awards

Gold Medal

Project Title

BuildGuard: Blockchain Oracle AI for Instant Secure Compliance

Winning Lingnan Faculty and Research Staff

• Prof Wu Liupengfei, Assistant Professor of the Division of Industrial Data Science at the School of Data Science

Project description

BuildGuard acts like a super-smart, tamper-proof inspector for remote manufacturing. It uses AI trained by experts and secure blockchain data feeds to check quality and compliance in real time, replacing slow manual reviews with instant, trustworthy, digital audits. The system transmits different types of production data securely to the blockchain ledger, where AI performs real-time analysis to ensure data cannot be altered. Through human-AI collaboration, BuildGuard transforms manual inspections into instant digital processes, solving the problems of low data trustworthiness and slow inspection speeds to support large-scale remote manufacturing while staying compliant with the latest regulations.

Key Features:

• Real-time Compliance Review

An AI core trained in specific domains analyses multi-modal production data on-site, autonomously conducting expert-level compliance checks.

• Blockchain Anti-tampering Mechanism

Integrated with blockchain oracle technology, inspection data is securely transmitted to the blockchain ledger, ensuring audit results are authentic and immutable.

• Accelerated Review Process

Converts time-consuming manual compliance reviews into instant digital workflows, significantly improving remote manufacturing efficiency and allowing high-volume production.

Awards

Gold Medal

Project Title

IDShield: AI Behavioural Fingerprinting for Anonymous Dialogue Decode

Winning Lingnan Faculty and Research Staff

• Prof Shen Jiaxing,Assistant Professor of the Division of Artificial Intelligence of the School of Data Science and Programme Director of the Master of Science in Data Science Programme
• Mr Wang Wenxuan, PhD student in the Division of Artificial Intelligence at the School of Data Science
• Ms Liu Zirui, PhD student in the Division of Artificial Intelligence of the School of Data Science
• Mr Kou Haoxuan, PhD student in the Division of Artificial Intelligence of the School of Data Science

Project description

In the digital era, anonymous mode does not mean true anonymity. Even when users are not logged in, use guest mode, or hide their IP addresses through a VPN, AI conversations may still retain traceable behavioural fingerprints, such as content preference, writing style, interaction patterns, and personality traits. These signals may create a risk of cross-session linkage, identity inference, and privacy non-compliance.

IDShield is designed to close this gap. Before content enters the AI platform, it detects and removes behavioural signals that could lead to re-identification. In doing so, it helps companies build safer, more verifiable, and regulation-compliant AI customer service and conversational environments, while preserving the meaning and usability of the interaction. 

Key Features:

• Behavioural Fingerprint Detection and De-identification

The system identifies behavioural cues that might reveal a user’s identity from multiple dimensions, including content, writing style, interaction patterns, and personality traits, thereby reducing the risks of cross-session linkage and re-identification.

• Proactive Anonymisation Protection

Unlike solutions that only mask accounts or IP addresses, this system processes identifiable features before content is sent to AI models, which closes the gap where users can still be recognised in anonymous mode, while preserving as much semantic meaning and business value as possible.

• AI Compliance Verification and Auditing

The system generates compliance verification reports, allowing companies to assess whether their AI conversational systems fulfil relevant requirements under GDPR Recital 26 and the EU AI Act, reducing regulatory risk and potential exposure to penalties.

• Stronger Trust and Better Business Outcomes

Through verifiable anonymisation protection, companies can not only strengthen privacy safeguards, but also improve user trust, platform adoption, and overall service.

Awards

Gold Medal

Project Title

Solid-Safe Lithium Battery: Puncture-Proof, 500Wh/kg, Compatible with All Lines

Winning Lingnan Faculty and Research Staff

• Prof Chen Ze, Assistant Professor of the Wu Jieh Yee School of Interdisciplinary Studies
• Prof Guo Ying, Assistant Professor of the Wu Jieh Yee School of Interdisciplinary Studies

Project description

This innovative all-solid-state lithium battery does not contain flammable liquid electrolytes, and offers enhanced safety and stability. It can be deployed to high-performance drones, electric vehicles, and large-scale grid energy storage systems that require extra safety. Currently, the battery has an energy density exceeding 500 Wh/kg, weighs nearly 40 per cent less than traditional power banks under equivalent conditions, and demonstrates significantly improved efficiency, charging from 20 per cent to 80 per cent in just 20 minutes.

Key Features:

• Exceptional Safety and Stability

The team has validated its stability through extreme tests such as needle penetration, overcharge and over-discharge, axe splitting, and fire exposure. 

• Enhanced Energy Density

Benefiting from the integration of “anode-free technology” that further reduces the weight and volume of the battery, its energy density has surged to 500 Wh/kg, placing it at the forefront of the current market. In the future, this advancement could potentially extend the range of electric vehicles to over 800 kilometres.

Awards

Gold Medal

Project Title

StorySketcher: A Multi-Agent AI System for Child-Led Co-Creative Narratives

Winning Lingnan Faculty and Research Staff

• Prof Xu Xian, Assistant Professor of the Department of Digital Arts and Creative Industries

Project description

This platform uses AI-guided questions to turn children's drawings into complete and unique stories, stimulating their narrative creativity, logical thinking, and language skills, and making the creative process fun.

Key Features:

• Enquiry-Based Inspiration

Through AI questioning and drawings, children are shown how to extend storylines they have created, and gradually construct a complete narrative.

• Child-Friendly Design for Co-Creation

The interface is suitable for independent operation by children, and also supports parental guidance and companionship.

Awards

Silver Medal     

Project Title

Face Fortress: Adaptive All-In-One AI for Digital ID Safety

Winning Lingnan Faculty and Research Staff

• Prof Sam Kwong Tak-wu, Associate Vice-President (Strategic Research), Dean of the School of Graduate Studies and J.K. Lee Chair Professor of Computational Intelligence
• Dr Guo Haifeng, Postdoctoral Fellow of the Division of Artificial Intelligence of the School of Data Science
• Prof Wang Shiqi, Professor of the Department of Computer Science at the City University of Hong Kong
• Mr Ou Fuzhao, Doctor of Philosophy (Student) of the Department of Computer Science at the City University of Hong Kong

Project description

This innovative system integrates deepfake detection and image quality filtering into a single verification pipeline. It first automatically assesses key quality indicators of facial images, such as clarity and lighting conditions, and then applies advanced algorithms to detect whether the image contains synthetically generated or manipulated facial content before proceeding to face recognition. The system is suitable for identity verification and public safety applications where accuracy and reliability are paramount.

Key Features:

• Dual-Layer Protection

Deepfake detection and quality filtering are combined within a unified workflow, eliminating the need to connect multiple systems, and considerably augmenting operational efficiency and reliability.

• Image Quality Screening

The system automatically evaluates uploaded facial images based on clarity, lighting, angle, and other quality indicators, filtering out blurred or non-compliant images to ensure high verification accuracy.

• Designed for High-Stake Scenarios

Delivers trusted verification results in accuracy-critical environments, such as opening a bank account, border control clearance, and other security-sensitive applications.

Awards

Silver Medal     

Project Title

Hydrogel Smart Window: Cool Without Power

Winning Lingnan Faculty and Research Staff

• Prof Chen Xi, Dean and Chair Professor of the Interdisciplinary Studies of the Wu Jieh Yee School of Interdisciplinary Studies
• Prof Ke Yujie, Assistant Professor of the Wu Jieh Yee School of Interdisciplinary Studies
• Mr Zhao Yu, PhD student at the Wu Jieh Yee School of Interdisciplinary Studies

Project description

This is a mass-producible hydrogel smart window that autonomously regulates indoor temperatures without electricity. The technology uses a water-retentive thermochromic hydrogel combined with low-emissivity film to filter solar heat radiation, cooling indoor spaces and reducing energy consumption by up to 34 per cent.

Key Features:

• Autonomous Regulation

Relying solely on the physical properties of the materials, it senses temperature changes autonomously, and adjusts dynamically in real time, for heat insulation in summer and thermal retention in winter.

• Significant Energy Saving and Carbon Reduction

Compared to traditional building glass, this can reduce indoor cooling energy consumption by up to 34 per cent, improving environmental sustainability.

Awards

Silver Medal     

Project Title

Smarter Traffic, Safer Roads: AIoT and Real-Time GeoAI Innovations for Urban Transport

Winning Lingnan Faculty and Research Staff

• Prof Paulina Wong Pui-yun, Head and Associate Professor (Presidential Early Career Scholar) of the Division of Science

Project description

This system is designed to tackle road congestion and improve road safety in Hong Kong. It combines AIoT hardware, cloud-based software platforms, and a GeoAI model to process and analyse dynamic traffic data as well as to forecast traffic conditions for the next hour. At the core of the solution are two actionable outputs: the Traffic Congestion Index (TCI) and the Traffic Accident Risk Index (TARI). These indices provide near-real-time information to drivers and fleet managers on future road conditions and give higher-risk driving contexts, improving safety, fleet management, and traffic congestion.

The project is funded by the Smart Traffic Fund (Project Ref: PSRI/56/2212/PR), and is currently being trialled on about 150 minibuses in Hong Kong.

Awards

Bronze Medal 

Project Title

DynaGuard: The Smart Brain for Dynamic System Monitoring and Diagnostics

Winning Lingnan Faculty and Research Staff

• Prof S. Joe Qin, President and Wai Kee Kau Chair Professor of Data Science
• Ms Chen Shumei, PhD student of the Division of Artificial Intelligence at the School of Data Science

Project description

This system works by learning from the normal operational data of machinery and equipment. When any subtle anomaly deviating from the norm occurs in a complex engineering system, it can instantly identify and issue an early warning, while conducting an in-depth diagnosis of the root cause. This allows engineers to intervene and address issues before a failure occurs, and also to pinpoint the source of the problem, improving operational safety and efficiency.

Key Features:

• Precise monitoring and real-time diagnosis

Through AI-driven autonomous learning of normal system dynamics, the technology improves the accuracy of anomaly detection, identifying hidden issues that traditional monitoring methods often overlook. It also diagnoses the type of anomaly, enabling an intelligent process of detection coupled with immediate diagnosis.

• Cross-platform applicability

Suitable for a variety of complex dynamic systems such as industrial manufacturing and healthcare. In future, it should be able to expand to smart city applications such as transportation networks and big data in financial systems, increasing diagnostic efficiency and precision.

Awards

Bronze Medal 

Project Title

Generative AI Assessment System (GAAS): A Game-Changer for Timely Student Feedback

Winning Lingnan Faculty and Research Staff

• Prof Frankie Lam King-sun, Director of the Teaching and Learning Centre
• Dr Rosiah Ho Wing, Project Consultant of the Teaching and Learning Centre

Project description

The Generative AI Assessment System (GAAS) is an innovative student-centred teaching tool that combines cutting-edge AI technologies such as Retrieval-Augmented Generation and Chain of Thought. It provides students with immediate and targeted feedback on their assignments, which saves teachers time marking work, allowing them to focus more on teaching, while also enabling students to track their own progress, strengthening learning effectiveness and experience.

Key Features:

• Automated Real-Time Feedback

Uses AI technology to provide students with immediate and targeted learning feedback so as to improve the effectiveness and experience of assignments, including graduation and service-learning projects. It also assists teachers to optimise assessment methods, thereby improving overall learning outcomes.

• Privacy Protection

All cases submitted for AI evaluation are automatically de-identified, meaning student names and ID numbers are not displayed. All data collected by the system will be permanently deleted after use.

• Driving Personalised Learning

GAAS accurately identifies the strengths and weaknesses in student assignments, and provides specific, actionable suggestions for improvement. Students can instantly see their personal learning progress, and teachers get a comprehensive view of students' learning trajectories, thus achieving the goal of using teaching assessment to promote student growth.

Awards

Bronze Medal  

Project Title

TransLab: AI-powered Translation Learning and Evaluation Suite

Winning Lingnan Faculty and Research Staff

• Prof George Chan Chi-yu, Assistant Professor of Practice of the Department of Translation

Project description

TransLab is an AI-powered educational platform specifically designed for translation and language learning. By integrating pedagogical principles into the learning process, providing context-specific feedback, and offering visual explanations for its evaluations, it significantly improves translation, learning and assessment, establishing a new teaching benchmark for efficient translation and language training.

The platform includes two main modules that both assist students in learning and support teachers in assessment, filling a gap in the existing market. The Learner Module (Transmuse) is designed for students and anyone learning to translate, offering a two-stage learning process from comprehension to improvement. It first helps users understand the source text properly, and then to consciously improve the quality of the translation. The Instructor Module (TransEval AI) is designed for teachers and evaluators, analysing translations using multiple criteria, highlighting error types and their causes, and providing specific revision suggestions based on context.

Key Features:

• Automated Assessment Process

By automating marking, it makes teachers more efficient, identifies error types and their causes, and offers specific revision suggestions based on the text's context, allowing teachers to focus on teaching.

• Promotes Autonomous Learning

The system improves students' self-learning translation and language skills without teacher guidance.

Awards

Bronze Medal

Project Title

CineSim: Interactive Cinematography Lab

Winning Lingnan Faculty and Research Staff

• Dr Tobby Kan Shiu-tao, Senior Lecturer of the Department of Digital Arts and Creative Industries

Project description

This is 3D film production simulation educational software designed for beginners, students, and professionals in the film industry. By integrating immersive VR technology, it creates a virtual film set where users can experience realistic production scenarios and practise principal filmmaking skills, such as cinematography, lighting, camera movement, and storyboard design.

The system overcomes the limitations of physical equipment and venues, fostering creativity, conceptual understanding, and technical skills while providing a cost-effective alternative to expensive physical studios and filming equipment.

The full desktop version of CineSim has been released on the STEAM platform, while the VR version is scheduled for release in the second quarter of this year.

Key Features:

• Developed with Unreal Engine

Users can select realistic 3D scene templates from multiple presets such as a nostalgic Hong Kong-style bing sutt or a public housing estate, and adjust lighting, switch between day and night, modify camera angles, and direct character movements to learn and practise filmmaking techniques. They can also develop their own scenes quickly in CineSim to improve pre-production efficiency.

• Cost-Effective and Efficient

Students can experiment with different setups and generate multiple versions to compare within minutes, significantly reducing the time required for pre-production planning and communication.