* denotes 2023 Innovation Award Winner
Cybersecurity Embedded into Solid State Hard Drive FirmwareWe have embedded data protection within the firmware of the SSD. Features include ransomware detection/protection, file-level encryption, zero-trust file access and multi-factor authentication. We have recently added Pre-Boot Authentication (PBA) to provide additional security for users.
HiDeNN-MCA: Hierarchical Deep-learning Neural Network enabled Multiresolution Clustering Analysis Software Toolkit for Evaluation of Mechanical Properties and Performance of Composite StructuresHiDeNN-MCA is a data-driven physics-based multiscale modeling software tool for evaluating mechanical properties and performances of composite structures while fully considering the microstructural features. It is established based on the integration of data science with mechanistic principles discovered using deep-learning neural networks.
Logistics Process Intelligent System (LoPIS)The invention is a tailor-made solution to the logistics industry, analyzing process data, calculating equipment-centric and order-centric metrics to automatically determine chokepoints in conveyor systems. Taking advantage of the abundance of log data, it incorporates newly developed data mining and deep learning techniques to predict process performance and reduce inefficiencies.
ARGUS – Anomaly Recognition Graphical Universal SystemARGUS is a universal anomaly recognition system for product surface quality control in the manufacturing industry. It can automatically adapt to different kinds of products according to their physical appearance and use AI to detect surface anomalies with a little number of positive samples as training data without labeling.
AI-based Cross-material Structural Health Monitoring Using Ultrasonic WavesA universal ultrasonic wave-based structural health monitoring and inspection method empowered by AI technique, which can identify the health condition (intact or damaged) and locating damage position for different materials, including metallic and composite materials.
R&D of a Vision System based on Fusion of AI Convolutional Neural Network and Novel 79GHz Millimeter-Wave Radar for Autonomous VehicleFusion of on-road AI object detection (vehicle & pedestrian) and advanced 79GHz radar technology to enhance multiple object detection and provide necessary information (including precise object speed, distance, location, type and size) for autonomous vehicles to avoid collision with frontal objects and enhance road safety.
AI Based Autonomous Vehicle Control System for MinibusIntegrating AI adaptive drive, sensor fusion and AI decision-making in its algorithm, this AI Based Autonomous Vehicle Control System for Minibus enables full auto-piloting and supports Level 4 auto-driving, thus, improving the accuracy of detection and the analysis capabilities of autonomous vehicles, thereby better enhancing vehicle safety.
Autonomous Air-Ground Cooperative Tunnel InspectorA UGV-UAV-cooperated autonomous system is specialised in autonomously capturing images in GPS-denied environments. With the employment of precise positioning techniques, air-ground cooperation and sensor fusion, locations of robots can be identified with high accuracy. Meanwhile, regular inspection of tunnel environments based on vision can be achieved automatically.
AI-based Linearization Schemes for 5G/6G Fiber/Wireless SystemsWe propose to apply Artificial Intelligence (AI)-based schemes to high-power 5G and 6G (the sixth generation of mobile communications) systems, aiming to deal with their intrinsic nonlinearities. We demonstrate high performance enhancement in real 5G networks at low-complexity and remarkable generalization capability for ensuring high quality-of-experience for the end user.
Automating Rheumatoid Arthritis Assessment Via Deep LearningRheumatoid arthritis (RA) is a chronic, autoimmune disease that most commonly afflicts joints in the hands, wrists and knees. The described invention is a deep learning algorithm that automatically scores X-ray radiographs of RA patients' joints and accurately assesses the severity of joint damage.
Datalink-Enabled AI for Fire OptimizationNeoskye proposes a datalink-enabled firing optimization mechanism based on networked AI tasks, including sensor data collection, target recognition, classification, and selection, tracking the target to be fired, selecting the right weapon, firing the weapon system at the right time, and overlaying a real-time multirobot system model with multiple kill chains.
DataTwin: A Proactive Data Management SolutionOur technology reveals unusual data entries, and checks them with the user prior to submission through format checks, data matching and machine learning algorithms. DataTwin helps organizations to identify and fix user errors before they are in database. It helps with information sharing and interoperability among the agencies.
Artificial Intelligence to optimise complex manufacturing or quantum technology control systemsManufacturing or quantum technology systems often have a large number of controls, slight variations of which can significantly change the outcomes. This technology is an AI algorithm that can quickly find the most optimal combination of such control parameters in these complex systems without requiring manual trial and error.
Edge AI-empowered Robot for Autonomous In-pipe InspectionThis technology empowers the data processing and decision-making at the network edge to reduce the data transmission and latency. It provides the fundamental supports for the robots to work efficiently and effectively in some extreme environments such as underground or underwater pipelines.
Blockchain Enabled, Supply Chain Transparency for Defense Supply ChainsAn application of blockchain technology (distributed ledger) to create an secure, immutable record of the requirements, claims and evidence that make up a supply chain - enabling users with rapid, trusted information about the origin of products they buy, consume and use.
Santos - the only physics-based virtual human modelSantos is the only physics-based digital human model integrating AI and enabling predictive human motion and behavior. Santos is a high fidelity model of the human body and human motion enabling digital twin like behavior with variables such as height, weight, strength, fitness, flexibility, and sex.
Trust-as-a-Service (TaaS) solutionVeridat.io introduces a fundamental tool for digital commerce by generating a foundation of truth with which anonymous parties can come together. Our technology allows them to place trust in each other’s evidence. Our Trust-as-a-Service (TaaS) solution is able to certify the existence, persistence, and content of digital documents.