|Name of innovation
|Click to download PDF factsheet
|5G O-RAN Base Station technologies
|An emerging 5G network deployment and business model that enables open, virtualised, and fully interoperable RAN, bringing more options for mobile operators.
|Limited choice of network equipment, high CAPEX and low deployment flexibility of equipment. Network resources not efficiently utilised, affecting user experience.
|5G Spectrum Sharing Technology
|Sharing licensed and unlicensed bands enhances 5G spectrum utilisation and reduces spectrum licence & equipment costs. Together with ASTRI’s independent consultation services, these advantages are encouraging enterprise 5G infrastructure deployment.
|Wideband 5G services require the use of a large amount of spectrum. 5G licensed cellular bands are limited and expensive. The major costs of 5G services are the spectrum licensing fee and new equipment cost. Both are high. Industry attention is focused on the public network, with little attention to and support for private and enterprise network deployment.
|5G Standalone (SA) Core Network
|A cloud-native and container-based standalone (SA) 5G core network allowing reliable and secure connectivity to network services and high performance UPF with ground-breaking throughput.
|Reliable, secure, ultra-low latency, high throughput, higher deployment agility and flexibility are just a few features offered by a 5G Core. With a 5G Core, the mobile communication experience can be greatly enhanced, with better end-user experience, quicker service to market time, simplified network operations and improved network capabilities. Vertical and enterprise markets can benefit from lower costs and more flexible deployments, with options of end-to-end network slicing, edge computing, scaling and upgrades.
|End-to-end V2X System
|An End-to-end Multi-tier vehicle to everything system with low-latency and reliable communications, which can enhance road safety and traffic efficiency.
|Smart Mobility is crucial for road safety and traffic efficiency;
A fast, safe, and reliable network with low latency is required for connected vehicles, pedestrians, and road infrastructures to share road safety information;
A central management system is required to manage V2X devices and provide different services. However, it's usually deployed in a data center, which will increase communication latency.
|Smart Indoor and Outdoor Geographic Information System
|Seamless indoor and outdoor geographic information system accommodating indoor positioning and massive telemetry data.
|Lack of highly scalable and cost-effective indoor and outdoor map service for the HK industry. Typical mapping tools are weak in indoor capability and are not optimised for users’ requirements. Massive real-time telemetry spatial IoT data is not visualised geographically over a scalable system. BLE device management is insecure, costly and power-consuming in a large area.
|Base Station for Heterogeneous Network
|5G networks adopt a heterogeneous network (HetNet) architecture, where 4G/5G macrocells and 5G small cells are used to provide enhanced coverage and services. Low/Mid/High-frequency bands are used together.
|Enable 5G network with improved data throughput over existing 4G networks, and enhance coverage as well as the performance of 5G networks.
|Cloud-based 5G technology Cloud-native mobile network based on AWS
|Cloud-native enables flexible network architecture in order to meet various 5G use case requirements. ASTRI has set up a 5G cloud-native network in Hong Kong-based on the AWS cloud services. The latency between Huawei CPE and AWS server is 5ms.
|High mobile network Capex, ultra-Reliable and Low Latency requirement in 5G vertical applications, scalable network capacity and on-demand service.
|Perception and High Precision Positioning
|ASTRI perception and high-precision positioning technology provide consolidated indoor/outdoor environment vehicle/human positioning with low latency. It is an infrastructure technology that lays the cornerstone for C-V2X, smart-city, smart-mobility, and Automated Guided Vehicle (AGV) applications.
|Too costly to install high-precision positioning devices (e.g., RTK devices) on all vehicles/humans to obtain their accurate position. Impossible/hard to obtain an accurate position in indoor and urban areas due to lack of or poor Global Navigation Satellite System (GNSS) signal. There is no solution to obtain all objects (vehicles/humans) in an area for road safety/smart transportation purposes. No consolidated indoor/outdoor positioning solution.
|Self-Shutdown Li-ion Energy Storage Device for Explosion Prevention
|Hong Kong Applied Science and Technology Research Institute Company Limited (ASTRI) has developed an Anti-Explosion Li-ion energy storage technology with a build-in self-shutdown layer (SSL) design that safeguards energy storage devices from thermal runaway.
|As an effort to safeguard lithium-Ion based energy storage devices from potential fire hazards and explosion risks caused by overheating, ASTRI has developed a safety enhanced Li-ion energy storage technology with a built-in self-shutdown design to greatly mitigate the hazardous risks and yield a more consummate and effective solution to protect the environment, properties, and human lives.
|High Performance Medium-Range Wireless Charging Technology
|Hong Kong Applied Science and Technology Research Institute Company Limited (ASTRI) has developed a medium-range wireless power transmission technology platform providing a safe and high-performance multi-device charging solution for smart homes.
|As an effort to address the constraints of short distance, specific orientation, low charging power, device overheated, and only a single device supported for conventional wireless charging solutions, ASTRI has developed a wireless power transmission technologies with a higher power rating over longer transmission distance for multiple device charging, higher tolerance of misalignment, and is a safer charging solution with reduced overheat problem.
|Highly Efficient Power Converters for Data Centre and Telecommunication
|Hong Kong Applied Science and Technology Research Institute Company Limited (ASTRI) has developed a new electronics packaging technology platform with improved electrical, thermal and reliability performance for key power conversion modules in data centres and telecommunication equipment.
|As an effort to address the challenges of supplying ever-increasing computing power using less energy in a smaller space with mission-critical reliability, ASTR has developed a new technology platform using Vertical-Driver-GaN and Package-Embedded Inductor technologies to improve the power density and the coupled electromagnetic-thermal performance concurrently for the critical power conversion modules in the next-generation data centres and telecommunication equipment.
|Enhanced NB-IoT Technology for wide frequency bands and different operators
|Hong Kong Applied Science and Technology Research Institute Company Limited (ASTRI) has developed an NB-IoT technology supporting a wide range of frequency bands with an advanced radio frequency transceiver and Release 14 compatible baseband.
|To address the need of NB-IoT device to cover a wide range of frequency bands with a reduced chip area, lower power consumption and enhanced performance, ASTRI has developed an NB-IoT technology that adopts an advanced polar transmitter architecture with on-chip power amplifier and advanced baseband algorithms for low power consumption, small chip area, multiple frequency bands and enhanced sensitivity.
|High-capacity Li-ion Energy Storage Device for Disposable Medical Capsule
|Hong Kong Applied Science and Technology Research Institute Company Limited (ASTRI) has developed a micro-nano structured Si/C composite anode design encapsulates nano-size Si aggregate into a micro-size C-based conglomerate that enhances the specific capacity of the anode of Li-ion energy storage device for disposable medical capsule application.
|To address the higher power and/or energy density requirement for rising disposable medical capsule applications, ASTRI has developed a high-capacity rechargeable lithium-ion energy storage technology. ASTRI's design employed micro-nano structured silicon/carbon composite, which aids in nearly double the anode specific capacity (700 vs 370 mAh/g) of Li-ion energy storage device for disposable medical capsule applications.
|Smart Camera for home security
|Hong Kong Applied Science and Technology Research Institute Company Limited (ASTRI) has developed a customised deep neural network (DNN) accelerator for privacy and security protection embedded in a smart camera.
|Most cameras record video without a filter, which causes privacy concerns from the public. Furthermore, those records contain enormous raw data, which is hard to use in smart city applications. ASTRI has developed a customised DNN accelerator embedded in smart cameras for smart city applications that contain face detection and object recognition with privacy protection by blurring the face.
|DC Building Power Protection by Digital DC Circuit Breaker
|Hong Kong Applied Science and Technology Research Institute Company Limited (ASTRI) has developed a semiconductor-based digital ultra-fast DC circuit breaker for protecting the power source, load, and cables in the LVDC grid from damage in abnormal electric conditions.
|To address the short circuit failure is DC voltage that cannot be effectively protected by DC mechanical circuit breaker, ASTRI has developed a semiconductor-based digital ultra-fast DC circuit breaker for LVDC applications to protect the power source, load and cables from damage in abnormal electric conditions, such as short circuit and overload, in a “microsecond range” protection speed.
|AI for manufacturing
|An advanced AI model development solution for multiple manufacturing applications under the conditions of data-hungry and qualified AI expertise shortage.
|While AI plays a key role in manufacturing, its implementation relies on a huge amount of data and AI expertise. Expanding applications for AI creates a shortage of qualified workers, therefore, hindering companies from adopting AI to increase productivity. Obtaining sufficient high-quality data is expensive, time-consuming, and sometimes infeasible, especially for small batch production.
|Vision Guide Industry Robot for Random Bin Picking
|An embedded Intelligent 3D object recognition system for random bin picking.
|Traditional automation involves robots that faithfully carry out specific actions repeatedly based on programs meticulously crafted by highly skilled engineers. These robots are inflexible and ill-fitted for producing goods with a short product lifecycle. In a high-productivity manufacturing line, parts detection and posture estimation must be performed in real-time.
|ARVR Compatible 4K2K HMD
|World's first true 4K2K ARVR HMD technology opened a new window to view the world.
|In the era of information explosion, people need to process more and more data every day.
A smart mobile device such as a smartphone is unable to provide first perspective information.
The existing head mount display suffers from low resolution, a low field of view and low brightness. The user is unable to effectively obtain adequate information or unable to have a truly immersive experience.
|Integrated Smart Inspection and Identification System for Jewellery and Antique
|Fast and compact size jewellery inspection system capable of identifying a massive number of gemstones with an affordable cost.
|Traditional gemstone inspection relies on gemstone testing laboratories.
Most of the existing tools for testing are bulky and expensive and not cost-effective for jewellery vendors to set up in the retail shops.
The inspection process usually takes at least a week to complete. Service charges and labour costs are high.
|Driver Behaviour and Fleet Routing/Scheduling in Fleet Management
|ASTRI’s developed an intelligent platform to address driving safety and operational efficiency of a vehicle fleet. By analysing driving behavior as well as the driver’s actions at the wheel, dangerous driving habits on the road can be identified, while the driver’s behavior can also be monitored. Dangerous driving habits, such as getting the vehicle too close to vehicles or pedestrians in front, may lead to accidents. Furthermore, undesired driver behavior such as talking on the phone, dousing off or emotional outbursts may jeopardise all passengers’ safety, actions that would warrant an alert to the driver. Last but not least, the platform also helps the vehicle fleet in scheduling and routing, targeting couriers or special transportation services.
|The driving behavior of drivers, especially those of public transportation, is critical to safe driving.
The emotion and behavior of said drivers are also critical to the safety of all passengers.
Inefficient scheduling and routing would add costs to a fleet’s operations.
|AI Analytic Engine for providing tailor-made trainings targeting SEN students
|ASTRI has developed a revolutionary and cost-effective distance-learning system that allows parents and teachers to conduct SEN-related training anytime. With the assistance of the programme recommendation engine, the system can also help educational psychologists to speed up the assessment process and create a more efficient Individualised Educational Plan (IEP) for SEN students.
|The number of SEN students is increasing to over 56,000 in public mainstream schools in 2020 – 21.
There are insufficient teaching resources for SEN students, such as a shortage of therapists and limited space in classrooms.
One-on-one intensive training is an effective therapy approach but is too expensive and unaffordable for most families.
|IoT Platform for Smart Water
|ASTRI has developed 2 solutions.
Integrating the technologies of t-distributed Stochastic Neighbor Embedding (t-SNE) and Principal Component Analysis (PCA) to build the Smart Water Analysis System (the Analysis System), which performs data filtering that identifies anomalies in real-time.
LoRaWAN-based Hong Kong Wireless Smart Water Metering System Standard and its Reference Design (the System Standard) is to facilitate the retrofitting of smart metering systems to existing buildings.
|Traditionally SCADA data analysis and water meter reading are performed manually.
SCADA data from many devices have to be analysed.
Meter reading is done once every 4 months, and no abnormal alarms can be generated. WSD is introducing wired smart water metering into new buildings, but it is difficult to retrofit existing buildings.
|Smart Credit Assessment & Analytics for Micro, Small & Medium Enterprises Financing
|Traditional credit scoring relies on the financial information provided by the loan applicants, but MSME lacks sufficient credit history and readily available financial records. Banks may find it difficult to assess their creditworthiness. In view of this, ASTRI conducted a study and explored the use of financial technologies to develop an alternative credit scoring. It can produce credit scores for MSME based on both alternative and conventional data.
|The challenge of traditional credit scoring relies on the financial information provided by the loan applicants of micro, small & medium-sized enterprises (MSME), such as financial statements.
The challenge of MSME who may not have sufficient credit history and readily available financial records. Without such data and visibility of their business operations, it may be difficult for banks to assess their creditworthiness which impedes their business expansion.
The cost concern of Banks who find the assessment of creditworthiness for MSME not cost-effective.
|Applying privacy enhancing technologies to alternative credit scoring
|ASTRI has developed a software platform that supports the deployment of advanced model frameworks, privacy protection based on different privacy-enhancing technologies, functions for evaluating privacy control and performance of the models and integration with the infrastructure of the HKMA for managing credit data shared by data partners and banks.
|To facilitate sharing of insights from alternative data between banks and data partners while complying with the requirements of the privacy regulations in Hong Kong.
To develop alternative credit scoring models using federated learning.
To verify the performance of the alternative credit scoring models and monitor the privacy control during the process of federated learning.
To address the above, ASTRI developed a software platform that supports the deployment of advanced model frameworks, privacy protection based on different privacy-enhancing technologies, functions for evaluating privacy control and performance of the models, and integration with the infrastructure of the HKMA for managing credit data shared by data partners and banks.
|Trusted Cross-Boundary Cold Food Import Safety Management Platform
|ASTRI applies the Internet of Things (IoT) and Distributed Ledger based Blockchain technologies with smart contracts to automate processes, enforce trust between stakeholders and provide visibility, accountability and traceability in the food supply ecosystem.
|Perishable food has been a challenge for managing food safety. Also, the current workflow involves much manual operation and paperwork processing. There is no mechanism to monitor seal lock status and the temperature compliance of the consignment en-route.
|Intelligent Financial Document Analysis for Regulatory Compliancy
|ASTRI’s Financial Document Analysis (FDA) system provides an intelligent information cognition system to analyse automatically and extract key information from the financial documents and conduct summary generation and compliance checking. It can enhance regulatory technology and reduce Non-compliance costs and Manual checking processes.
|The financial sector needs to handle various kinds of complex formatted documents. They need to understand, analyse and extract specific data information for regulatory compliance checking report generation, disclosure requirements, etc. However, the existing manual checking process is slow, error-prone and labour-intensive.
|AI-aided Medical Image Analytics for
Early Cancer Screening
|ASTRI has been developing an AI-aided medical image analytics system for collaborative diagnosis, image management, lesion detection, and classification in pathology images and capsule endoscopy videos. This technology can help doctors improve diagnosis and work efficiency.
|Medical image analysis is very time-consuming for doctors.
Impossible for doctors to analyse all collected medical images.
Images with unobvious features are easily overlooked by doctors with inadequate experience.
Medical diagnosis is location-restricted.
|Generic Information Extraction Platform for Unstructured Documents
|ASTRI’s Generic Information Extraction Platform utilizes an OCR engine and various computer vision techniques to convert the text in documents to digital format and extract useful information based on the user’s requirements. The platform enables financial and government institutes to automate data entry of different documents, for example, bank statements, land registry documents and salary proof. Therefore, It can reduce typos and labor costs.
|The conventional approach requires manual data entry to transfer data from document to system.
It is time-consuming and may contain typos, especially when entering a vast amount of data into tables.
Service charges and labour costs are high.
To address the issues, ASTRI’s Generic Information Extraction Platform utilises an OCR engine and various computer vision techniques to convert the text in the document to digital format and extract useful information based on the user’s requirements. The platform enables financial and government institutes to automate data entry of bank statements, land registry documents, salary proof, etc.
|Open Banking API Portal Connecting TSP to Bank Systems
|ASTRI develops an open banking API bridge portal to support HKMA’s initiative. It provides a unified API interface to third-party service providers (TSP) to access various bank systems with the authorisation bank client. The portal maps the unified API with different APIs of different banks. The portal also serves as a platform for dynamically deploying Fintech applications. ASTRI also develops a plug-and-play interface for the dynamic deployment of TSP services to the API bridge portal.
|There is a high barrier to developing fintech applications for accessing accounts of different banks because the bank clients can only use applications provided by banks to access their bank accounts. Also, it deters the introduction of innovative Fintech applications by third parties.
|Artificial Intelligence Chatbot for mixed Cantonese, English and Mandarin markets
|Customisable chatbot solution is adaptable to different industry domains in mixed Cantonese, English and Mandarin environments.
|Most chatbot solutions on the market are not capable of recognising conversations with mixed use of Cantonese, English, and Mandarin. Most chatbot solutions offer limited customisation capabilities for users to update knowledgebases with domain-specific vocabularies. Customisable chatbots are not user-friendly for non-technical users.