Listing Success | Frost & Sullivan Supports Shenzhen Haiqing Zhiyuan Technology Co., Ltd. in Successfully Listing in Hong Kong (1392.HK)

During the Hong Kong listing process, Frost & Sullivan carried out the following tasks: helping the issuer accurately and objectively understand its position in the target market, using objective market data to identify, support, and highlight the issuer’s competitive advantages, assisting the issuer, investment banks, and other intermediaries in writing important sections of the prospectus (such as overview, competitive advantages and strategy, industry overview, business, etc.), helping the issuer communicate with the Stock Exchange and investors, and assisting investors in quickly understanding the market ecosystem and competitive landscape, as well as providing feedback on various industry-related issues from the Stock Exchange to the issuer.

Frost & Sullivan has always been a leader in helping companies go public in Hong Kong. According to LiveReport Big Data (statistical data as of March 31, 2026), over the past 36 months and 12 months, and from January to March 2026, Frost & Sullivan provided listing industry consulting services for 195, 101, and 30 Hong Kong IPO listings respectively, ranking first in terms of quantity. It has rich industry experience and communication skills with regulatory authorities, exchanges, investment and financing institutions, and related organizations.

• In 2025, the top five participants in China’s entire multi-spectral AI enterprise sector combined held approximately 8.0% market share (by revenue), with our company ranking first;

• Lightweight edge AI computing technology enables autonomous edge AI processing through optimized small models deployed directly on devices, and the multi-spectral AI large model can be added at the backend as needed;

• With proprietary technologies in the multi-spectral AI industry, including AI algorithms, algorithm chips, embedded systems, optical and safety engineering, the company can provide end-to-end capabilities from design to manufacturing;

• The multi-spectral AI large model platform supports the company’s multi-spectral AI large model services, enabling customers to process and analyze data from multi-spectral devices.

According to Frost & Sullivan reports, by revenue in 2025, the company:

• ranked first in China’s multi-spectral AI companies, with a market share of 3.3%;

• ranked fourth in China’s multi-spectral AI module market, with a market share of 3.8%;

• ranked first in China’s multi-spectral AI large model services market, with a market share of 23.0%;

• ranked fourth in China’s multi-spectral AI perception terminal market, with a market share of 0.7%.

Perceptual intelligence refers to technology that uses artificial intelligence, especially machine learning and computer vision, to simulate human perception abilities, allowing machines to perceive, understand, and respond to the external environment through sensors and other devices. Based on the classification of simulating human sensory systems, perceptual intelligence can be divided into visual, auditory, tactile, olfactory, and gustatory perception. According to Frost & Sullivan reports, the market size of China’s perceptual intelligence industry increased from RMB 200.1 billion in 2020 to RMB 337.0 billion in 2025, with a compound annual growth rate of 11.0%. In the future, with technological progress and deeper commercialization, the overall market growth rate is expected to increase further, and the market size is projected to reach RMB 695.2 billion in 2030, with a compound annual growth rate of 15.6%. In 2025, visual perception still dominated the perceptual intelligence market, accounting for 65.0% of the total market size.

Source: Frost & Sullivan Analysis

The pan-security and robotics markets are core downstream application areas of perceptual intelligence, as they require high reliability in complex environments. The pan-security market is a demand-driven market, covering diverse application scenarios such as fire safety, food safety, urban monitoring, and industrial hazard prevention. It combines advanced perception, intelligent decision-making, and cost-effective deployment to achieve real-time risk detection and rapid response in complex environments. Currently, the robotics field places higher demands on the ability to perceive the real world—potential application scenarios for perceptual intelligence.

Source: Frost & Sullivan Analysis

The multi-spectral AI industry market in China continued to grow. The Chinese multi-spectral AI market increased from RMB 6.3 billion in 2020 to RMB 20 billion in 2025, with a compound annual growth rate of 26.0%. Driven by technological progress, this trend is expected to accelerate further, and the market size is projected to reach RMB 79.4 billion in 2030, with a compound annual growth rate of 31.8%. The Chinese multi-spectral AI module industry market increased from RMB 2.5 billion in 2020 to RMB 5.5 billion in 2025, with a compound annual growth rate of 17.1%. Driven by technological progress, this trend is expected to accelerate, and the market size is projected to reach RMB 20.2 billion in 2030, with a compound annual growth rate of 29.7%.

The Chinese multi-spectral AI large model services market developed rapidly, increasing from RMB 300 million in 2020 to RMB 1.5 billion in 2025, with a compound annual growth rate of 38.0%. Driven by technological progress, this trend is expected to accelerate further, and the market size is projected to reach RMB 8.4 billion in 2030, with a compound annual growth rate of 41.1%.

Source: Frost & Sullivan Analysis

In 2025, by revenue, our company ranked first among Chinese multi-spectral AI companies, with a market share of 3.3%. The following chart shows the market share of the top five market participants.

Source: Frost & Sullivan Analysis

In 2025, by revenue, our company ranked fourth in China’s multi-spectral AI module market, with a market share of 3.8%. The following chart shows the market share of the top five market participants.

Source: Frost & Sullivan Analysis

According to Frost & Sullivan reports, in 2025, by revenue, our company ranked first in China’s multi-spectral AI large model services market, with a market share of 23.0%, demonstrating significant market leadership. The following chart shows the market share of the top five market participants.

Source: Frost & Sullivan Analysis

In 2025, our company ranked fourth in China’s multi-spectral AI perception terminal market, with a market share of 0.7%. The following chart shows the market share of the top five market participants.

Source: Frost & Sullivan Analysis

●Comprehensive technical capabilities will become a core competitive advantage

Multi-spectral AI has evolved from an early modular and segmented product model to a full-chain model with system integration capabilities. Companies with full-chain capabilities including embedded modules, large models, and integrated hardware terminals offer greater flexibility in delivery formats, a wider customer base, and more comprehensive technical solutions. With platform-based and integrated delivery, they will have unique advantages in the future industry.

●Miniaturization of computing power

The key lies in achieving better application results with lower costs and less computing power requirements. For example, leading companies optimize underlying C language and task decomposition capabilities, removing redundant images to avoid unnecessary calculations, thereby reducing computing power needs.

●Sensor integration

Improved sensor integration and optical stack technology allow ultraviolet, infrared, and other visible spectrum data to be integrated into compact, low-power modules, enabling widespread deployment of multi-spectral perception in consumer electronics, industry, and automotive terminals.

●Edge AI computing architecture

Progress in edge AI computing architecture—including lightweight neural networks, edge computing, and high-performance DSP/NPU designs—enables real-time fusion of multi-spectral data, significantly improving performance in low-light imaging, material identification, and safety monitoring scenarios.

●Increasing customization to adapt to complex industrial scenarios

In the multi-spectral AI industry, demand for customized solutions is growing, making customized services a mainstream trend. In 2024, 55% of multi-spectral AI large model applications used customized delivery models, mainly targeting large enterprises and other types of users.

●Technical Barriers

This industry integrates multiple disciplines such as optics, electronics, software development, automatic control, digital image processing, artificial intelligence, pattern recognition, mechanical design, and manufacturing, making it a typical technology-intensive industry. Entering this industry requires not only the above-mentioned professional knowledge and practical experience but also a high-level technical talent and management team. It also relies on a large number of researchers to drive technological upgrades and skilled workers to continuously improve product quality. Additionally, technological breakthroughs require long-term and continuous investment in research and development. Currently, only leading companies in the industry possess strong independent innovation capabilities and can quickly respond to market changes, develop high-quality products, and gain market advantages, thus creating high technical barriers for new entrants.

●Talent Barriers

This industry involves a wide range of disciplines, including basic disciplines such as mechanics, optoelectronics, automation, and computing. Therefore, companies need a large number of researchers with complex backgrounds to achieve multi-disciplinary integration. Moreover, due to the highly customized nature of products, researchers must conduct feasibility assessments and develop process plans. The entire production process involves clear professional divisions, complex technology, and strong system coordination. Custom production requires high talent requirements in technology research and development, after-sales service, and marketing. Related talent training and team building require long-term accumulation, thus creating high talent barriers.

●Financial Barriers

This industry is financial and technology-intensive, and products fall under high-tech categories. To meet customers' requirements for accuracy and quality, products need continuous technological iteration and rely on substantial investment in research and development and funding. At the same time, large-scale production requires a large number of testing and processing equipment, and a complete sales and service network requires significant financial investment, thus creating high financial barriers.

●Data Barriers

Multi-spectral AI data includes not only regular image data but also raw spectral information collected in different bands. Such data collection is costly and requires specific hardware and on-site deployment, as well as specialized annotation teams to build scene semantic levels. Additionally, the high-dimensional nature of spectral data (hundreds of dimensions or more) makes data cleaning and enhancement more complex. Long-term operating enterprises build exclusive data asset pools deeply linked to model iteration through real-world data collection and feedback mechanisms, thus creating significant data entry barriers.