Author: MIL-OSI Publisher

Source: Republic of China Taiwan

According to the statistics, from January to April 2026, 796 foreign direct investment (FDI) projects with a total approved amount of US$ 7,308,500,000 were recorded. This represents a 23.99% increase in the number of cases, and a 158.08% increase in the FDI amount compared with the same period in 2025.

Regarding inward investment from Mainland China, 6 cases with a total approved amount of US$1,098,000 were recorded from January to April 2026. This reflects no change in the number of cases and a 98.86% decrease in the investment amount compared with the same period in 2025.

In terms of Taiwan’s outbound investment (excluding Mainland China), 207 projects were registered from January to April 2026, with a total amount of US$ 34,084,559,000. This represents a 15.85% decrease in the number of cases, but a 152.84% increase in the investment amount, compared with the same period in 2025.

As for Taiwan’s outward investment to Mainland China, 74 applications were approved from January to April 2026, marking a 27.59% increase compared with the same period in 2025. The approved investment amount totaled US$ 262,987,000, representing a 39.21% decline compared with the same period in 2025.

Source: Republic of China Taiwan

MIRDC has received the prestigious Bronze Award at the 2026 Edison Awards for its groundbreaking innovation, AI-Driven Fastener Forming Quality Monitoring.

The award-winning technology is transforming fastener manufacturing by integrating Smart Machine Box (SMB) sensors and AI algorithms to monitor forming force, predict defects, and safeguard process data through an encryption scheme that renders critical data features unrecognizable. The solution significantly enhances product quality, improves energy efficiency, and strengthens the global competitiveness of the fastener industry.

Taiwan’s fastener industry faces a number of challenges, including frequent mold changes, unstable forming forces, and fragmented production data. To address these issues, MIRDC has developed “AI-Driven Fastener Forming Quality Monitoring.” The system learns from a small number of high-quality samples, automatically fine-tunes parameters after mold changes, and tracks energy consumption and carbon emissions in real time. Chaos-based encryption is also incorporated to protect critical process data.

The system introduces three major innovations. First, it combines real-time raw material quality monitoring with AI-driven automatic parameter adjustment, reducing the defect rate from 10% to 0%. The system also predicts equipment health to prevent unexpected downtime and production interruptions, enabling zero-defect manufacturing. Second, the technology integrates the Manufacturing Execution System (MES) with smart meters to monitor machine utilization, power consumption, and carbon emissions, allowing precise management of production efficiency and carbon performance. Third, multiple chaotic systems and synchronization algorithms are applied to ensure the secure and complete transmission of critical production parameters. The encryption mechanism makes data difficult to identify, intercept, or decrypt, while reducing computational resource consumption by approximately 60%.

Overall, the technology enables zero-defect manufacturing and is already supporting key suppliers to Germany’s high-speed rail systems and leading global automakers. Going forward, the solution will be extended to other manufacturing processes, including die casting and metal stamping, enabling cross-industry replication and scaling.

Source: Republic of China Taiwan

MIRDC has received a prestigious Bronze Award at 2026 Edison Awards for its groundbreaking innovation, Next-Gen SHT, an advanced intelligent heat-treatment system designed for cast material.

As global demand grows for high-performance and net-zero manufacturing, traditional austempering processes still rely heavily on fixed parameters and manual experience, often resulting in repeated trial-and-error, lengthy development cycles, and inconsistent product quality. To address these challenges, MIRDC has developed the first Next-Gen SHT, integrating AI vision, adaptive learning, and intelligent alloy design. The system enables real-time adjustment of heat-treatment parameters based on material variation and performance needs.

The system utilizes high-precision AI vision technology to rapidly analyze the nodularity and microstructure of ductile iron with 95% accuracy, improving inspection efficiency by three times. The analyzed data further enables the AI module to continuously self-learn and automatically optimize alloy design and process parameters, maintaining heat-treatment error margin with plus-minus5%.

Furthermore, the system integrates three core modules: a high-temperature furnace, a low-temperature salt bath, and a filtration unit. These modules maintain stable furnace temperatures, preserve microstructure integrity, and achieve 80% recovery rate for treatment media.

Products processed by Next-Gen SHT demonstrate significantly enhanced performance, with tensile strength increased to 1,300 MPa and elongation reaching 13%. At the same time, the technology shortens development time by 40%, reduces process energy consumption by 71%, and lowers production costs by 66%.

This technology has already been applied to key components for green energy systems and agricultural machinery equipment. In the future, it will be further expanded into high-end applications in the fields of energy, smart electromechanical systems, and transportation vehicles.

Source: Republic of China Taiwan

MIRDC has received a prestigious Silver Award at 2026 Edison Awards for its groundbreaking innovation, and the innovative technology is ThermoMind-Smart Energy-efficiency Combustion System. The award-winning technology operates with optimal energy efficiency, and it features self-optimizing combustion control, high temperature precision, and heat exchange efficiency over 95%.

In Taiwan, traditional combustion furnaces rely heavily on operator experience for control and often suffer from high energy consumption and significant heat loss, resulting in substantial carbon emissions. To address these challenges, MIRDC has developed ThermoMind, a smart energy-efficient combustion system.

ThermoMind surpasses traditional manual control limits by achieving 72% waste-heat recovery, over 65% heat-exchange efficiency, and temperature precision within plus-minus3 Degrees Celsius. Its cloud-based diagnostics deliver real-time alerts and adaptive control to help prevent downtime, while the modular architecture enables rapid deployment across a wide range of furnace types and industrial processes.

To support the global transition toward greener manufacturing, the system is compatible with green fuels such as hydrogen, helping manufacturers reduce operating costs while improving sustainability performance. Since deployment, ThermoMind has helped reduce more than 111,554 tons of Carbon dioxide emissions.

In other words, ThermoMind transforms traditional combustion furnaces into self-learning smart systems that regenerate waste heat, boost efficiency, and support green fuels. It has been deployed in 23 industrial sites, including Ying Chien and Taiwell Aluminum, driving industries toward a net-zero future.

Source: Republic of China Taiwan

MIRDC has received a prestigious Silver Award at 2026 Edison Awards for its groundbreaking innovation, and the innovative technology is MagicABC. The award-winning technology is an AI-Guided Interactive Speech-Language Therapy System, and it aims to help children with speech delays.

Taiwan is facing a shortage of Chinese speech-language pathologists, which limits the amount of clinical therapy available for children with speech delays. To address this challenge, MIRDC has developed MagicABC, an innovative speech-language therapy software designed to help alleviate the shortage of Chinese speech-language pathologists in Taiwan.

Most speech-language therapy software offers one-way therapy. MagicABC integrates Chinese clinical data into Speech-Language Pathologist-Agent, embedding a virtual speech therapist into games. Through two-way communication and instant feedback, it breaks the limits of traditional one-way therapy.

Furthermore, the software introduces three major innovations. First, it integrates interactive guidance, speech recognition, and text-to-speech to create a dynamic feedback cueing mechanism that simulates common therapeutic prompting techniques based on user responses. Second, it features AI-driven therapeutic guidance. The software combines clinical and general corpora with in-context learning to dynamically adjust dialogues according to context and learning progress. Third, the software applies Natural Language Processing (NLP) to evaluate language abilities in semantics, syntax, and pragmatic, and then generating personalized learning reports.

For applications, MIRDC has partnered with more than 10 medical and educational institutions, and 81% of cases show improvements in language comprehension and verbal expression. In the future, the software will expand into applications such as aphasia rehabilitation, special education, home-based rehabilitation, and cross-language training, with the goal of building a sustainable intelligent speech therapy ecosystem.

Source: Hong Kong Information Services

The Home Affairs Department today appealed to eligible registered electors to vote this Sunday in the Sha Lo Tung Cheung Uk, Kam Tsin Wai Tsuen, Wa Shan Tsuen and Yung Shue Ha Indigenous Inhabitant Representative Elections, and the Lin Fa Tei Resident Representative Election.

Nineteen Rural Representative vacancies are open in the by-election, and 19 valid nominations were received during the nomination period.

Two candidates are running for Indigenous Inhabitant Representative posts on each of the following committees: the Tai Po Rural Committee in Sha Lo Tung Cheung Uk; the Pat Heung Rural Committee in Kam Tsin Wai Tsuen; the Sheung Shui District Rural Committee in Wa Shan Tsuen; and the Lamma Island (South) Rural Committee in Yung Shue Ha. Two candidates are also running for a Resident Representative post on the Pat Heung Rural Committee in Lin Fa Tei.

Polling will be held for these five vacancies.

The other nine candidates were returned uncontested. These are the candidates to be Indigenous Inhabitant Representatives of Ha Yeung, Tai Po Mei, Kat O, Ma Wan Main Street and San Tong Po, and the candidates to be Resident Representatives of Sheung Ling Pei, Shek Pik San Tsuen, Wo Liu Hang and Sha Tau Kok Market (West Lower).

No valid nominations were received for the remaining five vacancies, including the Indigenous Inhabitant Representative vacancy at Kwan Mun Hau, and the Resident Representative vacancies at Shan Pui Chung Hau (I), Shui Tin Tsuen, Tai Po and Tap Mun.

Polling will be held from noon to 7pm.

The department reminded electors to bring along their original identity documents, or other specific alternative documents, to the polling station as specified in the polling notice.

Source: Hong Kong Information Services

The Housing Bureau announced that the latest quarterly Composite Waiting Time for Subsidised Rental Housing is 4.7 years, marking the lowest level in over eight years. Secretary for Housing Winnie Ho noted it is also the first time that the waiting time has dropped below five years.

“We are confident that within this term of Government, the Composite Waiting Time would be capped at five years.”

As the Government strives to fulfil low-income families’ housing needs, about 9,650 Light Public Housing (LPH) units have been fully occupied as at the first quarter of 2026, while 20,150 units will be completed in the remainder of 2026, gradually moving towards the target of completing the construction of 30,000 LPH units by 2027-28.

Better living

Ms Zhuang and her son, who moved into an LPH unit in Kai Tak last year, are two of the citizens who benefit from the LPH policy. She previously rented a subdivided flat in To Kwa Wan, where conditions and hygiene were far from ideal. With the extra space at the new home, they can spend more time together and are in a happier living situation.

“Since moving into LPH, my son’s concentration has improved, and his attention deficit hyperactivity disorder symptoms have eased thanks to the better living environment.”

Total solution

In the first quarter of 2026, about 8,400 general applicants have been housed to public rental housing (PRH) or LPH. Around 47% of them were housed to LPH units. Notably, the waiting time for PRH units in urban districts was about six years, while that for LPH units in the New Territories was only about two years.

Ms Ho said LPH provides a total solution for those stuck in poor living conditions, and urged families in need to consider moving into the extended urban areas, such as Tuen Mun, as years of poor living conditions may take a great toll on the young.

“The four-year difference means if you have a child in primary one, a six-year-old, within four years, he is already a 10-year-old.

“Four years in a very poor living condition without the confidence to interact or talk to other people will affect him for the rest of his life.

“When I see children move into LPH, they have more open space to play. They have this neighbourhood, friendship.”

Moreover, she pointed out that moving into LPH, as compared with renting subdivided units, is a financially advantageous option for families in need.

“On average, each family can save $5,500 per month.”

For less financial burden, a better family relationship and improved living conditions, households in need should look at LPH as a much better option, the housing chief advised.

Source: Hong Kong Government special administrative region

Following is the speech by the Financial Secretary, Mr Paul Chan, at the Hong Kong Polytechnic University (PolyU) Doctor of Business Administration (DBA) 30th Anniversary Dinner today (May 14):

Professor Jinguang Teng (President of PolyU, Professor Teng Jinguang), Professor Edwin Cheng (Dean of the Faculty of Business of PolyU), DBA alumni and students, distinguished guests, ladies and gentlemen,

Good evening. It is a pleasure to be here tonight to celebrate the 30th anniversary of PolyU’s DBA programme. Founded in 1996, it was the first of its kind in Hong Kong – and it remains to this day the largest DBA community in this city and across the Mainland.

Building a community of scholar-leaders over three decades is no small feat. My warmest congratulations to everyone who has contributed to this achievement.

A DBA programme is an endeavour to bridge rigorous research with real-world practice, translating ideas into decisions that move businesses and societies forward. Tonight’s theme – “From Ideas to Impact” – perfectly captures the spirit that has defined this programme from the very beginning.

I am particularly pleased to see that tonight’s panel will explore how AI and Web3 are turning ideas into business and societal impact, with a focus spanning supply chain management, digital asset infrastructure and robotics. The breadth of that agenda reflects the true scope of the transformation before us – and it is a transformation we must confront with the utmost attention.

Technology and business transformation

Let’s take a moment to reflect on how technology and innovation have transformed business and commerce.

Back to 1996 – the year this programme was born. The Internet was in its infancy. A mobile phone was the size of a water bottle. Data was a scarce resource. Business strategy was often about optimising within known boundaries: how to produce more efficiently, distribute more widely, price more competitively.

What we have witnessed since then is not incremental improvement, but the rewriting of the grammar of commerce. Consider these: we used to buy a CD; now we stream. We used to purchase software; now we subscribe. We used to own the things we relied on; but increasingly, ownership has evolved into a greater emphasis on getting access. Products have become platforms.

And value is no longer created solely within individual companies; it emerges across ecosystems, where capital, talent and data flow across organisational boundaries supported by interoperable systems, digitalisation and sophisticated logistics.

Profound challenges of AI era

Today, we have entered the age of AI – a moment more far-reaching and more disruptive than anything that came before. The challenge is no longer simply about adopting new tools. It is about recalibrating the way we think – and doing so with urgency.

First, AI is compressing the distance between insight and action. A strategist might once have spent months analysing market data before moving, but an AI agent today monitors market sentiment in real time, simulates scenarios and recommends courses of action or even has them executed – all at once. The bottleneck is no longer information, but our ability to direct machines with precision, and to verify and act on what they present.

Second, AI is redrawing the architecture of value chains. Middle layers that were justified by their ability to co-ordinate and process, are being compressed. At the same time, new intermediary roles are emerging – data curators, model trainers, compliance architects. I suspect many of you are quietly asking yourselves: in the new organisational structures that are taking shape, where do I stand? Will I still sit at a node that adds value?

For Hong Kong, these shifts carry particular weight. As an international financial centre and a super connector, our competitive edge has long rested on intermediation – in capital, information, markets and talent. AI will challenge each of these roles. But it also gives us the opportunity to reinvent ourselves and move forward.

Opportunities we must seize together

This brings me to the core message I want to leave with you tonight. The opportunities ahead are enormous – but they will not materialise on their own. They require purposeful, co-ordinated action among the Government, business and academia. I want to highlight three areas where our collective leadership can make a decisive difference.

First, the digital workforce. The conversation about AI and jobs tends to focus on displacement – on what AI will take away. But that is only one side of the story. The more important question is what AI makes possible. Businesses need digital employees to scale what they do. Employees need digital assistants to keep them focused on what they do best. AI agents, if deployed well, can become a genuine extension of both. They can absorb the repetitive, the routine and the administrative, so that human energy can be directed to where it matters most: judgement, relationships and creative problem-solving.

But none of this works if the gains stay at the top. When AI doubles what a team can produce, the people who made that possible need to feel it – in their salaries, in their roles, in their prospects. That is how a virtuous cycle begins. Companies that share the gains through new skills, new responsibilities and better rewards will not just outperform their peers. They will be the ones that attract and keep the best people.

Second, the human-machine relationship. In the future, much of the information that companies produce – press releases, annual reports, regulatory filings, even research papers – may be written, processed and read primarily by AI systems, with humans as the secondary audience. When machines become the primary intermediaries of information between organisations, what matters is not just what you communicate, but whether your systems can communicate with theirs. Interoperability becomes the new baseline requirement.

But alongside that shift comes an equally important truth: distinctly human skills – curation, verification, ethical judgement – become more valuable, not less, precisely because they are what machines cannot be trusted to do alone. The question is whether our institutions, our training systems and our own habits can adapt quickly enough to ride on this development.

Third, AI literacy. We must treat AI literacy as a societal imperative. In this year’s Budget, we outlined significant investment in computing infrastructure, the AI+ strategy and providing AI Training for All. But government effort alone will not be enough. We need a coalition of employers, educational institutions, professional bodies, business chambers and trade unions to build AI capability at every level of the workforce.

The history of technology tells us that the fruits of innovation tend to concentrate before they spread – accruing first to innovators and those who own capital, and reaching workers and communities only much later. The steam era followed that pattern. So did the early Internet. AI presents us with a rare opportunity to do better because it can empower even the ordinary people.

When a small business owner can deploy a digital agent to handle customer enquiries around the clock; when a nurse can spend less time on paperwork and more time at the bedside; when a teacher can tailor lessons to every student in the classroom – that is the moment productivity becomes prosperity. That is the transformation towards which we should build.

Concluding remarks

Ladies and gentlemen, 30 years ago, PolyU had the foresight to create a programme that would turn experienced executives into scholar-leaders – people who think critically, research rigorously and act decisively. That foresight is worth more today than it has ever been. The world does not lack ideas. What it often needs is the capacity to turn ideas into impact – to move from concept to execution, from pilot to scale, and from individual achievement to collective progress.

Tonight is therefore both a celebration and a call to action. A celebration of what this programme and its remarkable community have built across three decades; and a call to all of us to ensure that the transformation underway is pursued with ambition, guided by responsibility and designed to benefit the many, not just the few.

That is the future worth building – together.

Thank you very much.

Source: Hong Kong Government special administrative region

Source: Hong Kong Government special administrative region