![](\"https://www.getech.cn/files/edit/20251017/vlbPyr4QE6J10KCxUNYa6Hhispa8p88F.jpg\")
In the semiconductor manufacturing field, data-driven precise decision-making is becoming the core competitiveness for improving yield and reducing costs. However, traditional data analysis systems often face three challenges, making it difficult to quickly extract the value of large amounts of industrial data.
- High access threshold: Relies on professional SQL writing or complex BI tool operations, blocking many business personnel from accessing data.
- Long analysis cycle: From data extraction, cleaning to modeling analysis, a complete yield root cause analysis often takes several days, causing severe decision lag.
- Difficult knowledge inheritance: The analysis experience and ideas of senior engineers are hard to accumulate and reuse, forming a dilemma of 'knowledge exists with the person, experience is lost when the person leaves'.
Getech's latest GT Insights product is designed to break this dilemma. As a conversational business intelligence analysis system based on natural language processing and large model technology, GT Insights is redefining data interaction methods in semiconductor factories. Unlike traditional BI tools, GT Insights supports users in directly obtaining required data analysis and visualization results through natural language dialogue, reducing the threshold and cycle of data analysis to an unprecedented level through AI workflows, realizing the vision of 'data democratization'.
The core breakthrough of GT Insights lies in achieving a natural interaction mode of 'conversation as analysis'. Engineers only need to ask simple natural language questions, such as 'Show the yield trend of Product A in the third quarter', and the system can automatically understand the intent, completing the entire process from semantic parsing, SQL generation, data query to visualization presentation in the background, lowering the threshold for data analysis.
Its underlying technical architecture revolves around the semantic modeling layer, innovatively solving many challenges of LLMs directly processing physical databases in industrial scenarios. By building a semantic layer of business indicators and dimensions, the system transforms complex physical table structures into business-friendly data models, enabling LLMs to generate accurate semantic SQL, which is then converted into executable physical SQL through the semantic layer, achieving accurate understanding of business terms and generating reliable analysis results. This design not only significantly improves query accuracy but also makes the entire generation process verifiable and traceable, laying a solid foundation for subsequent intelligent analysis.
From a technical architecture perspective, GT Insights adopts a distributed deployment scheme, supports cluster expansion, ensuring high availability and stability of the system. Its core components include the semantic modeling layer, multi-agent framework, LLM integration layer, and StarRocks distributed data warehouse, together forming a complete industrial-grade intelligent analysis platform.
![](\"https://www.getech.cn/files/edit/20251017/uTorrdo4zzZdXJO9zaFZJhFBFBwoqcWf.jpg\")
Yield analysis is the most disruptive application scenario of GT Insights. In traditional models, a low yield root cause analysis often requires engineers to manually extract data across different systems, try various statistical methods for data collection, organization, and multi-dimensional cross-analysis, taking several days. GT Insights, through preset AI workflows, enables the system to automatically execute the complete analysis chain, compressing this process to the minute level.
For example, when a user asks 'Analyze the reasons for low yield of Product A last week', the system automatically triggers a multi-step analysis process: first, intent recognition to lock the product and time range; then, execute Daily BIN analysis to locate abnormal BINs and time points; then, initiate HeatMap overlay analysis to identify defect distribution patterns; then, through Commonality analysis, trace back to suspicious machines; finally, generate a full-chain analysis report. The entire process requires no manual intervention, automatically generating a structured analysis report with clear direction and complete evidence chain including abnormal BINs and problem machines within minutes.
At the functional level, GT Insights demonstrates comprehensive data processing capabilities. From basic chart generation and multi-dimensional drill-down analysis to complex yield root cause localization and SPC statistical process control, the system can quickly respond through conversational interaction. Its unique Dashboard auto-generation function allows users to simply describe requirements in natural language, and the system intelligently arranges analysis charts to form a complete monitoring view.
![](\"https://www.getech.cn/files/edit/20251017/iN4YnC7o11WtRcEr0yphDhcymjzxa19q.png\")
The value of GT Insights goes beyond breakthroughs in single-point capabilities; it is an important component of Getech's CIM AI Foundation, developed based on Getech's Octopus AI Agent Platform 3.0, and can collaborate with intelligent 'digital employees' such as Planner Agent and Yield Analysis Agent through the AI Agent framework. It moves from passive response to active planning of analysis paths, calling third-party tools (e.g., SPC software, yield analysis plugins), forming closed-loop analysis decision-making capabilities.
![](\"https://www.getech.cn/files/edit/20251017/ZCPL7Le7TXD2udNDcdj6mmwMYZnfQwXR.jpg\")
At the system integration level, GT Insights offers 'driver' and 'co-driver' modes. It can be used as an independent analysis platform directly or embedded as a plugin into existing business systems like MES and YMS, allowing users to summon the AI assistant for auxiliary analysis while operating business interfaces, achieving a seamless experience of 'business operation as analysis'.
For semiconductor CIM systems, the introduction of GT Insights marks an important evolution from 'process automation' to 'intelligent decision-making'. Traditional CIM systems mainly focus on production process automation and data collection, while GT Insights endows these massive data with intelligent analysis capabilities, enabling manufacturing execution systems to truly possess intelligent features of self-perception, self-analysis, and self-decision.
Currently, Getech GT Insights has been successfully deployed in multiple semiconductor manufacturing enterprises, helping customers achieve significant results in yield improvement, equipment efficiency optimization, and production cycle shortening. For example, a semiconductor packaging and testing company originally faced issues like long yield analysis cycles, difficulty in accumulating expert experience, and data silos across multiple systems. By deploying Getech GT Insights, the company built full-chain analysis capabilities covering yield data, traceability information, and AOI images, significantly improving product yield and greatly reducing system operation costs and employee training costs.
With the continuous maturation of industrial large model technology, conversational BI represented by GT Insights is becoming an important direction for digital transformation in the semiconductor CIM field. Getech combines advanced AI technology with deep industry knowledge to provide a practical path for intelligent upgrading in the semiconductor industry, helping Chinese semiconductor manufacturing move towards higher efficiency, precision, and intelligence. In the future, GT Insights will continue to deepen specialized capabilities in specific semiconductor scenarios, expand solutions for more vertical industries, promote industrial intelligence from 'usable' to 'easy to use', and inject new momentum into the digital transformation of manufacturing.
From September 28 to October 4, the Bangkok Climate Action Week was held for the first time in Bangkok, Thailand. Gechuang Dongzhi's Southeast Asia Chief Architect Deng Xiaojin (Simon) was invited to attend and deliver a keynote speech. As a key participant in green and low-carbon development, Gechuang Dongzhi fully demonstrated its profound expertise in smart manufacturing and energy management. Its AI-powered energy and carbon management solution, originating from the semiconductor manufacturing industry, can be innovatively applied to the green building sector and provides crucial support for the development of green buildings in Southeast Asia.
![](\"https://www.getech.cn/files/edit/20251017/pLt1Z4JbuuBuJljnw3S1KKJJYEEyBQYA.png\")
The Bangkok Climate Action Week is a climate week centered on cities, targeting the public and multiple stakeholders, aiming to mobilize societal forces and promote green transformation and climate resilience building in Southeast Asia. The event brings together city governors, businesses, youth, and communities through diverse scenarios such as exhibitions, performances, seminars, policy labs, and community actions, bringing climate issues into daily life and urban decision-making. The event advocates 'whole-of-society participation,' focuses on replicable low-carbon pathways and climate resilience practices, and emphasizes the empowerment of local forces.
As global climate challenges become increasingly severe, Southeast Asian countries are accelerating their green transformation. Thailand, as a regional energy hub, is speeding up its green energy transition, especially in green buildings, proposing 'Towards Net-Zero Energy Buildings' and pointing out that cooling equipment is a key area for energy efficiency improvement. In response, the Thai government has introduced the BEC Building Energy Code, TREES Green Building Certification System, and strengthened air conditioning energy efficiency standards (TIS 2134-2565) to promote energy saving and carbon reduction in the building and industrial sectors.
As a leading enterprise in China's industrial intelligence industry, Gechuang Dongzhi deeply integrates manufacturing genes with AI technology to create an 'Energy + Carbon Emissions + AI' smart energy-carbon management platform. It monitors and manages the real-time energy consumption of electricity, cooling, heating, and other multi-energy sources in buildings, while using AI algorithms and cloud computing technology to focus on the 'major energy consumers' in industrial and commercial buildings—HVAC systems—achieving online monitoring, energy consumption prediction and simulation, and real-time optimization strategy control for HVAC equipment and systems.
![](\"https://www.getech.cn/files/edit/20251017/JftIPkapa2P16kc0t4XjQKGSS06bNTzr.jpg\")
By organically combining AI intelligent algorithms such as Light GBM and Random Forest with mechanistic models built based on the physical characteristics of equipment/systems, it achieves real-time control of equipment like chillers, fans, and pumps, ensuring the entire system always operates at optimal energy efficiency, maximizing energy utilization efficiency, reducing carbon emissions, and ultimately helping building facilities become green and low-carbon.
The platform has capabilities for monitoring, control, reporting, analysis, predictive diagnostics, and AI energy-saving optimization, encompassing the basic functions of BAS/FMCS systems, significantly improving the flexibility and intelligence of HVAC optimization, with an average annual energy saving rate of 20%.
- Visual energy usage monitoring: Intuitively displays the real-time operating status and key parameters of systems and equipment;
- AI predictive diagnostics: AI algorithms achieve predictive fault diagnosis and provide resolution suggestions;
- AI strategy energy-saving intelligent control: AI optimization algorithms automatically control, achieving efficient energy saving and unattended operation.
At the same time, it features full functionality, low cost, fast delivery, and high availability. It is compatible with different brand hosts and has wide application scenarios. It includes a self-developed multi-protocol adaptation engine, with high versatility and compatibility, supporting over 95% of common industrial protocols on the market, making application scenarios broader, achieving plug-and-play equipment, and solving protocol isolation issues. With advanced algorithm capabilities and self-control integration, it incorporates dynamic optimization algorithms, enabling system modeling based on physical laws for the entire system (water side + air side), performing professional and comprehensive chiller performance prediction, and achieving 8760-hour system simulation with real-time dynamic optimization.
In addition, Gechuang Dongzhi can provide end-to-end industrial intelligent solutions covering energy input, energy consumption, and carbon emissions, offering AI-powered comprehensive digital solutions for energy and carbon emissions for industrial, manufacturing, and commercial sectors:
- EMS Energy Management System: Panoramic monitoring, energy efficiency analysis, intelligent operation and maintenance
- CMS Carbon Management System: Carbon emission monitoring, carbon footprint accounting, emission reduction verification
- AI Algorithm Optimization Applications: Key energy-consuming equipment monitoring, energy consumption simulation and prediction, AI optimization strategy operation
![](\"https://www.getech.cn/files/edit/20251017/IP2r6SVEhiyS40dJA0l1RVt3utp50MDz.jpg\")
On the communication platform of Bangkok Climate Week, Gechuang Dongzhi demonstrated the practical path of AI + green technology. Under the guidance of the dual carbon goals, Gechuang Dongzhi will continue to use AI and digital technologies to empower the manufacturing and construction industries in Southeast Asia to achieve green upgrades, working with customers to build an efficient, low-carbon, and sustainable future.
In the article 'Getech Stocker: Covering Semiconductor Full Process Storage, Promoting Yield & Efficiency Dual Acceleration', we mentioned that to cover the full process storage needs of semiconductors, Getech has built a complete Stocker product matrix of 'core storage + supporting cache', including core products such as wafer cassette storage equipment (Normal Stocker), tower cross-floor storage equipment (Tower Stocker), reticle cassette storage equipment (Reticle Stocker), nitrogen purge storage equipment (N2 Purge Stocker), etc., and supporting equipment such as overhead cache equipment (OHB), near equipment cache equipment (NTB), etc.
As a key role in the matrix focusing on the 'last meter near equipment', Getech NTB is an automated caching device customized for semiconductor production equipment (lithography, etching, deposition, etc.). Deployed next to the equipment, it fully utilizes vertical space to increase effective buffering, achieves efficient material transfer between 'OHT overhead hoist - production equipment', solves pain points in traditional logistics such as equipment waiting for materials, OHT congestion, and manual intervention contamination, thoroughly opening up the 'last meter' of semiconductor material automated flow.
![](\"https://www.getech.cn/files/edit/20251013/48YBaNOYmSdvwihL0E3TPlht58AqvpN1.jpg\")
In the semiconductor AMHS system, OHT overhead hoist handles 'cross-region long-distance transport', Stocker handles 'full-line centralized storage', and Getech NTB, with its three features of 'close distance, high frequency, fast switching', becomes the 'key link' connecting both to the equipment.
![](\"https://www.getech.cn/files/edit/20251013/ZbipPLWWLP5O191sVpYt5OB2uC0k7B8c.png\")
Flexible space deployment: Uses modular拼接 design, footprint only 1-2㎡, can be embedded in gaps of etching machine clusters, narrow channels beside testers, etc., without occupying core logistics pathways;
Second-level material handover: Distance to equipment Load Port only 1-3 meters, single Load/Unload time as fast as 10 seconds, much faster than OHT average response time, eliminating delays in logistics turnover, balancing rhythm differences between equipment.
Continuous material change capability: Robotic arm uses lightweight design, with servo motor precise control, fastest single Load/Unload time ≤10 seconds, can complete at least 3 continuous Load/Unload tasks in 1 minute, perfectly adapting to high-frequency change needs of testers and sorters;
Local autonomous decision: Built with Getech self-developed control system, can directly receive equipment EAP commands, no need to rely on MCS system scheduling, avoiding response delays from network congestion.
E84 protocol is the core communication standard for equipment (EQ) and external devices (e.g., OHT/NTB) to achieve data interaction. In traditional mode, E84 switching has pain points like manual reconfiguration, long switch cycles, and error-proneness. Getech NTB innovatively develops E84 one-key quick switch function, no need to change hardware wiring, via NTB local touchscreen or remote MCS system, click 'Switch to OHT'/'Switch to NTB' to trigger, complete mode switch within 10 minutes, providing key support for flexible production lines.
Getech NTB is not just a simple 'temporary shelf', but through 'pre-stocking, automatic handover, buffer scheduling', brings value like equipment utilization improvement and AMHS efficiency optimization to semiconductor production lines.
In traditional production mode, equipment utilization is typically only 75%-80% due to waiting for OHT delivery, Load Port full停机, etc. Getech NTB achieves efficiency breakthrough through two mechanisms:
Pre-cache materials to be processed: MCS system issues transport tasks to OHTC based on equipment production rhythm, OHTC schedules overhead hoist to send material boxes to NTB, equipment can immediately receive materials after completing current work order, improving response timeliness, avoiding 'equipment waits for materials';
Temporarily store processed materials: Processed wafer carriers from equipment can be temporarily stored in NTB, waiting for OHT idle time to集中转运 to Stocker, avoiding equipment停机 risks from Load Port full.
If all materials are directly handled by OHT to equipment, it leads to frequent OHT scheduling, track congestion (especially when multiple equipment call for materials simultaneously). Getech NTB balances AMHS load through 'buffer adjustment':
Reduce OHT无效往返: OHT can send material boxes for multiple equipment to corresponding NTB in one go, then集中转运 processed material boxes to Stocker, single task coverage expands, OHT空跑率大幅降低;
Alleviate track congestion: Avoid OHT频繁启停,排队等待 in equipment areas,提升整体AMHS吞吐量.
As one of the few domestic companies with 'AMHS overall solution + core equipment self-research' capability, Getech NTB's value is not only in solving the single pain point of 'equipment waits for materials', but also in串联起AMHS全链路协同 through 'small device' – from efficient scheduling of OHT overhead hoist, to centralized storage of Stocker, to continuous production of equipment, Getech NTB becomes a key link in打通自动化物流闭环.
As of now, Getech NTB has累计出货20+ units, with 95% applied in domestic leading 12-inch advanced process wafer fabs. In the future, Getech will continue to focus on 'technology self-controlled,场景深度适配, service efficient response' as the core, strengthen and improve the 'AI+CIM+AMHS' overall solution,助力半导体工厂智能化升级.
![](\"https://www.getech.cn/files/edit/20251013/hFAt705W7EyIEyOctGqCaKL4x13XMQK5.jpg\")
On October 11, the 2025 Optics Valley AI+ Industry Innovation Forum with the theme 'Wisdom Gathers in Optics Valley, Embarking on a New Journey' was grandly held in Wuhan. The industrial intelligence leader Getech was invited to deeply participate in the joint initiative release of 'Opinions on Deeply Implementing the AI+ Action', clarifying the direction for the landing and transformation of AI technology. It also showcased key achievements in the 'Optics Valley AI+ Industry Innovation Achievement Exhibition Area', jointly presenting the strong strength of the Optics Valley AI industry cluster with benchmark enterprises such as Peking University Wuhan Artificial Intelligence Research Institute and Huagong Technology.
![](\"https://www.getech.cn/files/edit/20251013/ILYSSLFU8pCiUmQRWcaKiK2Hw7N0az5R.jpg\")
This grand event was jointly organized by the Optics Valley Entrepreneurs Association and China Information and Communication Technology Group, hosted by FiberHome Telecommunication Technologies, and co-organized by many AI technology and application innovation benchmark enterprises and institutions in Optics Valley. It aims to gather wisdom from all parties, deeply explore the potential of AI, promote the deep integration of AI and industry, and empower high-quality development through innovative practices.
Based on over 40 years of know-how accumulation in the pan-semiconductor manufacturing industry where TCL is involved, Getech, with its integrated strategy of 'AI + Industrial Software + Intelligent Equipment', centrally showcased its 'AI + CIM + AMHS' integrated software and hardware solution that empowers semiconductor manufacturing. In terms of results, this solution has been implemented and delivered in 8-inch and 12-inch wafer fabs and packaging and testing plants, helping semiconductor factories achieve 100% full-process quality traceability, increase the utilization rate of key equipment by over 20%, and directly enhance production line capacity and economic benefits. It is noteworthy that the company has successfully delivered digital and intelligent projects in multiple stages such as plant construction, ramp-up, and expansion for dozens of leading Chinese semiconductor enterprises, with a customer satisfaction rate of 100%.
![](\"https://www.getech.cn/files/edit/20251013/jJ5ZSm6WBRrDE9YcSbOsJTqiLkW5KCUn.png\")
Solid delivery capability has won high market recognition. So far, Getech has successfully served nearly 300 pan-semiconductor customers, cumulatively delivered over 1,200 projects, with a customer repurchase rate as high as 70%, establishing Getech's industry-leading position as number one in market share in China's pan-semiconductor market. Behind every number is the best testimony of working side by side with customers to overcome challenges and create value together.
![](\"https://www.getech.cn/files/edit/20251013/s2Ive60HEdE5lh7bJSR2Q4iY2VczFUkS.jpg\")
This participation not only precisely echoes industrial innovation but more importantly, through deep exchanges and ecological co-construction with partners in Optics Valley and China's semiconductor industry chain, it actively promotes the intelligent leap of key technologies from independent controllability to industrial chain upgrading. In the future, Getech will accelerate technological innovation, create customer value, and together with industry peers, provide key support and clear examples for the high-quality development of Optics Valley and even China's semiconductor industry.
On October 22, the '2025 China Semiconductor Advanced Packaging and Testing Conference and 2025 China Semiconductor Wafer Manufacturing Conference', regarded as an industry development bellwether, will kick off in Kunshan, Jiangsu. The conference gathers global semiconductor experts, scholars, business leaders, and other stakeholders to discuss the core theme 'Wafer manufacturing is the foundation, advanced packaging is the breakthrough direction'. It will release new wafer manufacturing processes, Chiplet architecture, and other cutting-edge achievements, and address industry pain points like efficiency under equipment constraints and advanced packaging bottlenecks through forums, attracting over 500 leading companies for cooperation.
Yang Jun, head of Getech Dongzhi's advanced packaging and testing business, is invited to attend and will deliver a keynote speech titled 'From Automation to Autonomy: The New Generation Evolution Path of AI-Enabled Advanced Packaging CIM', sharing his practical experience and insights. Meanwhile, as a leading industrial intelligence enterprise focusing on semiconductor digital transformation, Getech Dongzhi will showcase advanced packaging and testing product solutions and results at booth A008. We look forward to your visit to grasp industry trends and explore future directions together!
![](\"https://www.getech.cn/files/edit/20251010/1kmZ1cprhK2YMYgAS6CU75SEID7wclbF.jpg\")
Today, the fourth TCL Employee Creative Competition public welfare project category final review results were announced. GETECH 3 public welfare projects, with their unique creativity, outstanding social value, and strong responsible spirit, stood out and were successfully selected as “TCL Excellent Public Welfare Projects”!
3 projects cover the fields of art education, sports public welfare, and environmental protection, demonstrating the multi-dimensional social responsibility and commitment of GETECH people. Subsequently, TCL Public Welfare Competition will provide these projects with resources such as funding, training, and communication, helping public welfare ideas move from blueprint to reality.
The project will go to Chiang Rai Province, Thailand, to support local Chinese schools for descendants of the expeditionary force. By offering traditional culture courses such as Chinese language, paper-cutting, and folk customs, it aims to alleviate the crisis of Chinese cultural inheritance among overseas Chinese youth, enhance their ethnic identity, and protect the cultural roots deep in their blood with practical actions.
![](\"https://www.getech.cn/files/edit/20251009/uk5MtDPW6yLkOnLD3GIkcqK1HtcjWT2q.png\")
The project will go deep into Hualong, Qinghai, to provide professional training camps for local basketball-loving youth. We not only donate supplies and teach basketball skills but also strive to sow the seeds of the Olympic spirit on the plateau, empowering growth through sports and addressing the shortcomings of physical education in remote areas.
![](\"https://www.getech.cn/files/edit/20251009/oaMAne5F0zfFJ1phWnT6WOA2eDtPFq06.png\")
Ecological Keba · Responsible Coexistence – Building a Green and Beautiful Qinghai Plateau Together
The project is rooted in Keba Village, Hualong, Qinghai, continuously promoting afforestation in high-altitude ecologically fragile areas. Aiming to restore the ecology of the Three Rivers Source and explore the \"ecology+\" development path, this project plans to plant about 750 saplings and add irrigation equipment, contributing GETECH's strength to the national “dual carbon” goals.
![](\"https://www.getech.cn/files/edit/20251009/J2UP1LXAqVP9a1xeyxL4WYAxWjCn0qev.jpg\")
The heart of public welfare starts with small steps. GETECH people always uphold the belief of contributing to society with professionalism and responsibility. From the company's mountain area used book donation activity this year to the current 3 public welfare projects, though the forms vary, they together constitute our most vivid interpretation of “responsibility”. The three project leaders also have profound thoughts:
“Public welfare is never an individual's impulsive dedication, but a long-term cause undertaken by like-minded partners with a sense of responsibility. Public welfare is not charity, but shared responsibility; not the pity of bystanders, but the responsibility of actors. For sixteen years, with this heart, I have practiced this journey. I hope to join hands with all GETECH members to go all out on the public welfare path and move forward resolutely!”
“Responsibility is a promise to oneself and those around, and public welfare is sharing this heart with more people. Both make me feel a deep sense of connection, that my existence can make the world a little better.”
“Doing public welfare with a responsible heart is the natural extension of responsibility from the workplace to society; practicing responsibility through public welfare actions is the solid implementation of goodwill from concept to action.“
3 projects are a vivid practice of the “responsibility” culture in the public welfare field. This responsibility to society, the environment, and the future is precisely the force driving us forward.
GETECH's public welfare path, GETECH people keep moving. A new public welfare invitation is on the way. In 10月份, we will launch the company's public welfare project, looking forward to the participation of all partners!
GETECH established Getech Vision through strategic acquisition, enhancing AMHS (Automated Material Handling System) hardware-software integration capabilities, especially deep integration of AMHS with CIM system, and incorporating AI technology, further improving system efficiency and reliability. This article, from the perspective of Getech Vision, analyzes the logistics upgrade demands in the semiconductor industry, discusses the localization development of AMHS, and how AI technology drives AMHS intelligent upgrade.
\r\nAutomated Material Handling System (AMHS), as the 'logistics artery' of semiconductor factories, provides efficient and stable automatic material handling, significantly improving production efficiency, increasing product yield, and reducing human error rates, thus widely used in silicon wafer plants, wafer fabrication plants, packaging and testing plants, and other front-end and back-end semiconductor processes. In recent years, AMHS has accelerated localization, with AI technology playing a catalytic role, promoting domestic AMHS companies to seize the market in product intelligent upgrades.
\r\n![](\"https://www.getech.cn/files/edit/20251009/ZYk5sg0725Im43L1hTQM2GO4ka5AJ0FF.png\")
GETECH Technology Co., Ltd. (referred to as GETECH) is an industrial intelligent solutions provider strategically incubated by TCL. In the semiconductor industry, it provides integrated 'AI+CIM+AMHS' hardware-software solutions for semiconductor factories. In 2024, through strategic acquisition, it established a new subsidiary—Shanghai Getech Vision Intelligent Technology Co., Ltd. (referred to as Getech Vision). Currently, Getech Vision has built a comprehensive AMHS product system including OHT (Overhead Hoist Transport), Stocker wafer storage library, MCS material control system, OHTC intelligent OHT control system, STKC control system, etc. In this feature, Getech Vision's relevant personnel detailed the logistics upgrade demands in the semiconductor industry and the innovative development of AMHS technology.
\r\nGetech Vision: In recent years, the development path of China's semiconductor industry has significantly shifted from comprehensive catch-up to a new stage of technological breakthroughs and全面提升 manufacturing capabilities. This is mainly driven by market demand. China's rapid development in industries such as new energy vehicles, photovoltaics, 5G, and the rapid iteration and application of AI technology in various industries have directly led to a surge in chip demand. Influenced by geopolitics, China continues to increase investment in the semiconductor industry, with new factory construction growth far exceeding the global average.
\r\nFrom the manufacturing perspective, in mature processes, 28nm and above chips have achieved mass production in China and are widely used, with technology and产业链趋于成熟; third-generation semiconductors (such as silicon carbide, gallium nitride) are developing rapidly, with China at the forefront in this field; advanced packaging and testing technologies (such as Chiplet, 3D IC) have also made significant progress in China, with leading companies ranking top in the global market and steadily expanding market share. However, in advanced processes, although there are technological breakthroughs in 14nm and below chips in China, large-scale mass production is not yet possible due to equipment procurement restrictions.
\r\nFrom the logistics perspective, as semiconductor manufacturing processes become more advanced, wafers evolve to larger sizes, from early 6-inch wafer plants to 8-inch wafer plants, and now 12-inch wafer plants have become mainstream. The number of processes in production lines has multiplied, and the throughput of key equipment has rapidly increased, making manual material handling between production line equipment a bottleneck restricting production efficiency and quality. The urgent need for logistics automation upgrades in enterprises has led to significant growth in AMHS market demand. However, compared to the substantial improvement in manufacturing capabilities, the development of logistics automation in China's semiconductor industry is relatively slow.
\r\nGetech Vision: The most prominent change in the domestic AMHS market is the rapid rise in demand and increasingly widespread application. AMHS has become an indispensable logistics support for 12-inch wafer plants. With the increase in the number of newly built 12-inch wafer plants in China, the demand for AMHS as their standard配置 has surged. At the same time, domestic AMHS has made significant technological progress, gradually capturing the market in areas such as 8-inch wafer plants and advanced packaging and testing, and也开始批量应用 in 12-inch mature processes (≥28nm). It can be said that both demand and supply sides are driving force, promoting further market scale growth. Relevant institutions predict that China's AMHS market will exceed 16 billion yuan in 2025, with a compound annual growth rate of over 12%.
\r\nThe second significant change is that customers require AMHS to upgrade from basic handling to intelligentization. Early AMHS only completed the搬运 of wafer boxes (FOUP) from point A to point B. Now, it is required that AMHS deeply integrate with MES/CIM systems to achieve dynamic path optimization and real-time scheduling. This is because as factory production scales expand and the number of equipment increases, the efficiency requirements for AMHS further improve, while also needing to enhance system safety and reliability. This can be achieved through全流程 system simulation, equipment layout optimization, track optimization design, intelligent scheduling, intelligent path planning, intelligent traffic management, OHT speed, and other aspects. For example, Getech Vision introduces AI algorithms to predict congestion, improving OHT system scheduling efficiency by over 20%.
\r\nAdditionally, domestic semiconductor companies have raised other requirements for AMHS software, such as improving user-friendliness. Foreign companies, due to earlier establishment and technological R&D, have become僵化 in technology, such as older software systems and unfriendly human-machine interfaces. Domestic manufacturers excel in this aspect.
\r\nCurrently, wafer plants for 14nm and below have extremely stringent requirements for AMHS stability and safety, and this market is still mainly dominated by Japanese companies. Achieving high-precision, high-reliability scheduling for over a thousand OHTs and efficient coordination of ultra-large-scale equipment are key challenges that domestic AMHS companies urgently need to突破.
\r\nGetech Vision: Getech Vision's parent company—Getech East Wisdom is an AI-driven industrial intelligent solutions provider, offering 'AI+CIM+AMHS' integrated solutions for泛半导体 customers, covering manufacturing execution, equipment automation, quality management, energy and carbon management, logistics automation, digital supply chain, and other functions. AMHS is the core抓手 for achieving unmanned logistics in semiconductor factories and is regarded as a key part of the solution. To add hardware components on the basis of CIM solutions, in 2024, Getech East Wisdom strategically acquired Yunde Co., Ltd. and Jiangsu Ruixinku Intelligent Technology Co., Ltd., merging to establish a new subsidiary—Getech Vision. Through resource integration, it完善 the AMHS product system to better help semiconductor industry customers achieve automation, digitalization, and intelligent upgrades.
\r\n![](\"https://www.getech.cn/files/edit/20251009/JRGQI5t5GqQlwyKHOUtNxXkqd8cEIvNH.png\")
Currently, Getech Vision focuses on semiconductor industry AMHS solutions as its core business, while also engaging in泛半导体 warehousing automation and non-standard automation businesses. Getech Vision's AMHS方案 builds a 'software + hardware + algorithm'三位一体 system: in hardware, it covers core equipment such as OHT and Stocker wafer storage library; at the software level, with the material control system (MCS) as the中枢, it links real-time dispatching system (RTD) and real-time scheduling system (RTS) to achieve全流程 digital control; combined with our self-developed scheduling algorithms, it makes material handling more precise and efficient.
\r\n![](\"https://www.getech.cn/files/edit/20251009/gXzBUk3CasWhkDObbtkBn9MWA5lQzv3N.png\")
Getech Vision has established strong R&D teams domestically and internationally, achieving multiple technological breakthroughs, such as: delivered Tower Stocker and N2 Purge Stocker, filling the technological gap in domestic high-end storage equipment; self-developed wireless power supply system, breaking the foreign monopoly on this technology; self-developed OHT controller and OHT operating system, achieving powerful edge computing capabilities and multi-sensor fusion perception, making OHT not just passively execute tasks but complete tasks more intelligently, safely, and efficiently.
\r\nGetech Vision: Stocker brings benefits such as yield improvement, efficiency optimization, and cost reduction to production lines through全流程 automated storage, system协同 scheduling, high洁净 control, etc., becoming an accelerator for promoting semiconductor intelligent manufacturing. As one of the enterprises in the domestic Stocker field with 'earliest R&D, most complete varieties, and most shipments', Getech Vision Stocker has the following advantages:
\r\nFirst,全场景 equipment matrix, covering semiconductor全流程 intelligent storage. Getech Vision provides wafer box storage equipment (Normal Stocker), tower cross-floor storage equipment (Tower Stocker), mask box storage equipment (Reticle Stocker), nitrogen-filled storage equipment (N2 Purge Stocker), as well as supporting equipment such as overhead buffer (OHB), near-equipment buffer (NTB), etc. At the same time, it can be optionally configured with manual port/OHT port/AGV port/Auto Teaching, etc., through refined hardware-software module combinations and dedicated configuration tools, quickly responding to customer customized storage needs.
\r\nSecond,极速 delivery + 7*24 service, ensuring rapid production line deployment and stable operation. With deep semiconductor industry Know-How, strong supply chain support, professional project delivery teams, and installation teams, Getech Vision Stocker can achieve设备交货 in as fast as 2.5 months, and complete installation and调试交付 within 3 months,缩短 over 50% compared to international brands (average 6~8 months). At the same time, it establishes an all-weather service mechanism, with the fastest response to customer feedback within 2 hours and on-site maintenance within 4 hours.
\r\nThird, nearly a hundred units shipped, receiving consistent recognition from front-end and back-end customers. From ensuring front-end advanced process wafer storage safety to supporting back-end packaging and testing multi-scenario flexible production, Getech Vision Stocker, with its core advantages of 'technologically self-controllable,场景深度适配, service高效响应', has累计 shipped nearly a hundred units, gaining consistent recognition from front-end and back-end customers. In the future, Getech Vision will continue to optimize Stocker product performance (such as improving handling efficiency, positioning accuracy, and safety redundancy) to meet the demands of semiconductor production lines for scaling and automation upgrades.
\r\nFourth, CIM+AMHS深度联动, driving full factory automation. Getech Vision's parent company Getech East Wisdom provides 'AI+CIM+AMHS' integrated solutions, offering one-stop logistics automation upstream and downstream systems, including MES, RTS, MCS, RTD, OHT, Stocker, etc., avoiding multi-vendor equipment interface incompatibility and data不畅 issues; at the same time, through hardware-software integration and whole-factory control, achieving 'storage-handling-production'全链路 data打通, helping factories achieve unmanned and efficient operations.
\r\nGetech Vision: If AMHS is divided into software and hardware两大 parts, traditionally, software is the decision-making layer of the entire system, and hardware is the execution layer. For example: MCS decides to move the wafer box from point A to point B, OHTC receives the instruction and decides which OHT to dispatch and what route to take to move the wafer box from point A to point B; OHT receives OHTC's instruction and executes the handling operation.
\r\nGetech Vision enhances OHT's edge computing capability and fusion perception capability, and self-develops the OHT operating system, gradually decentralizing部分 decision-making authority to the OHT hardware layer, achieving central planning and remote decision-making, which can more efficiently and safely realize wafer box handling. The OHT operating system has also become an important part of AMHS software. For example,就像 OHT driving on tracks,以前 controlled by traffic lights (OHTC) at intersections; after OHT becomes intelligent, it can exchange information and autonomously decide the travel path,就像 having a traffic police实时指挥 at the intersection, making the system more agile and efficient.
\r\nGetech Vision: AI technology undoubtedly has a profound impact on various industries. For semiconductor manufacturing CIM systems, it is undergoing a transition from 'experience-driven' to 'data-driven + AI-empowered',重构 from 'process automation' to 'intelligent decision-making中枢'. Similarly, AMHS also has many important decision-making issues. Adopting AI technology for intelligent upgrades can achieve optimization in the following aspects:
\r\nFirst, improve AMHS efficiency through全流程 system simulation and intelligent scheduling algorithms. Based on the analysis of多年 operational data, it can achieve whole-factory process simulation and AMHS system design optimization, including equipment layout optimization, OHT track laying optimization, OHT scheduling optimization, etc., to improve efficiency, achieving the same tasks with fewer OHTs;再加上 intelligent scheduling algorithms, further reducing the number of OHTs and congestion, thereby全面提升 AMHS efficiency. This is Getech Vision's unique advantage. Compared to AMHS suppliers who only consider the logistics环节, we can truly consider and打通 all aspects of production and logistics from top to bottom based on the CIM system, achieving whole-factory design optimization and efficiency improvement.
\r\nFor example, semiconductor factories typically have thousands of equipment. Through RTD and RTS, if it is found that the quality indicators of certain equipment are declining, task allocation to those equipment can be gradually reduced; predictive path planning can also be performed for the OHT system, for example, if it is found that a certain OHT has low operational efficiency, it will not be assigned more tasks.
\r\nSecond, predictive maintenance, improving system reliability. Traditional system maintenance involves定期 equipment inspection. Based on AI+OHTC system实时 monitoring the status of key components such as OHT motors, reducers, various sensors, etc., combined with Agent, it can achieve significant improvement in fault warning accuracy and fault analysis processing efficiency, thereby better ensuring stable and reliable system operation.
\r\nThird, AI-assisted fault diagnosis. In the past, after a fault occurred, it relied on experienced experts to judge and solve the problem. With AI-assisted diagnosis, even novices can quickly resolve faults.
\r\nFourth, equipment intelligentization. In addition to changes at the software level, the aforementioned equipment such as OHT will also become intelligent. Because只有 software intelligentization without hardware intelligentization, we believe AMHS intelligentization is difficult to go far.
\r\nGetech Vision: Getech East Wisdom has深耕 industrial AI for many years and foresees that AI will bring major innovations to AMHS, so it took the lead in establishing a joint industrial AI laboratory with the University of Hong Kong. The laboratory will deeply integrate the University of Hong Kong's AI research advantages with Getech East Wisdom's industrial AI落地 case experience in serving advanced manufacturing, conducting research on the direction of 'industrial AI technology innovation and practical application'.
\r\nThe first R&D project currently推进 by the laboratory, the AMHS intelligent material handling scheduling system, targets the industry pain points of domestic AMHS 'insufficient stability, low technological maturity', aiming to accelerate breakthroughs in OHT scheduling AI algorithm technology, achieve intelligent, efficient, and precise scheduling of ultra-large-scale OHT systems with over a thousand units, strongly promoting domestic substitution. The OHT intelligent scheduling model currently under R&D will achieve perception fusion of MES, MCS business data, and OHTC operational data, forming a highly business-aware AMHS solution, accelerating whole-factory efficiency optimization.
\r\n![](\"https://www.getech.cn/files/edit/20251009/TDKgR3Rd7BkH2hcFDG9EfWP3s6z0k8dK.png\")
AMHS intelligentization will be a long-term research topic, so we have leased a new office in Hong Kong Cyberport, where professors and scholars from the University of Hong Kong, and R&D engineers from Getech East Wisdom and Getech Vision jointly conduct research. The AMHS intelligent material handling scheduling system has been divided into multiple sub-topics and is being推进 in an orderly manner.
\r\nGetech Vision: Getech Vision, with its 'software + hardware + algorithm'三位一体 AMHS automated material handling solution, has累计 served over 50 semiconductor enterprises, including many industry leaders, covering front-end wafer manufacturing and back-end packaging and testing全场景.
\r\nFor example, in the front-end field, an 8-inch wafer plant of a military保密 unit faced material handling efficiency challenges under multiple equipment and complex process flows. We deployed a domestically self-developed 'MCS+OHT+track+OHB' system for them, allowing wafers to automatically flow between different processes, completely告别 manual handling errors and delays, laying a solid logistics foundation for stable and efficient production line operation. In a leading 12-inch wafer plant, the Stocker automated storage solution we delivered not only met high cleanliness and high safety requirements but also刷新 the industry record with '15-day delivery of the first equipment',背后 by our deep understanding of semiconductor processes and strong supply chain support.
\r\nIn the back-end packaging and testing field, we also have benchmark cases. In a leading packaging and testing enterprise,结合 its logistics digital upgrade needs, we provided a one-stop hardware-software solution including MCS, RTD, Stocker, building a whole-factory unmanned logistics system, becoming a model for large-scale automation application in the packaging and testing industry. It is worth mentioning that for special storage needs, our self-developed N2 Purge Stocker nitrogen-filled storage library can实时 maintain a class 100 ultra-clean environment, with抗震 design ≤0.5G, and can achieve 'lights-out operation' through automatic scanning and AGV linkage, eliminating risks from human intervention at the root, which is particularly critical in environmentally sensitive semiconductor manufacturing.
\r\nThus, Getech Vision's AMHS solution is not just an upgrade of 'handling tools' but also重构 the factory's logistics logic through digital means—both improving space utilization, allowing every inch of cleanroom space to maximize value; and ensuring absolute safety in material flow and storage, ultimately providing solid logistics support for semiconductor factories to reduce costs, increase efficiency, and improve quality.
\r\n*This article is authorized to be reprinted from the WeChat public account 'Logistics Technology and Application' (ID: LMH_56)
\r\nRecently, Suzhou Huaxing Optoelectronics Technology Co., Ltd. [hereinafter referred to as TCL Huaxing (Suzhou)] won the Ministry of Industry and Information Technology's certification for the '2025 Excellent Level Intelligent Manufacturing Demonstration Factory' with its 'AI-driven new display panel smart factory', indicating that it has reached the leading level in China. Entering TCL Huaxing (Suzhou)'s 540,000 square meter production plant, highly automated robotic arms and overhead transport tracks operate orderly, with few workers visible. The core supporting the factory's efficient operation is a smart operational system with self-decision and self-execution capabilities, a star intelligence model that accelerates product innovation from the source, and AI digital employees acting as the 'eyes' and 'brain' of quality inspection engineers.
![](\"https://www.getech.cn/files/edit/20251009/T7g8u4XNcX7d8wlg3ASIIiD1fv217sln.jpg\")
In the key step of micron-level defect inspection for factory products, AI digital employees operate through a closed-loop 'ADC+ADR' dual-system: the ADC Automatic Defect Classification System relies on AI visual technology to automatically identify and classify over 4,000 types of micron-level defects with 95% high accuracy; the ADR Automatic Defect Repair System, upon receiving instructions, drives high-precision tools to perform fully automatic repairs on repairable defects. This system replaces over 75% of traditional manual labor, upgrading the traditional 'human eye + experience' model to a precise and efficient new paradigm of 'AI + data'.
![](\"https://www.getech.cn/files/edit/20251009/Dmni89cs0LY24SEXRz5dHv1XllKF4IdU.jpg\")
Compared to traditional AOI equipment, which relies on manual re-judgment with misjudgment rates as high as 60%-70%, the digital employees based on the 'ADC+ADR' dual-system are driving the arrival of the 'super intelligent era' in industrial quality inspection. Behind all this is the deep application of GeChuang DongZhi's self-developed Tianshu AI ADC visual inspection system (hereinafter referred to as Tianshu AI). GeChuang DongZhi's Tianshu AI encapsulates the complex computations and operations involved in deep learning into ready-to-use productivity tools, bringing significant efficiency improvements, considerable cost savings, and precise defect identification capabilities through 'millisecond-level re-judgment', redefining 'quality inspection':
1) From 'mass manual re-judgment' to 'AI second-level judgment'. Traditional AOI automatic optical inspection equipment, which can improve initial inspection efficiency, still has misjudgment rates as high as 60%+, requiring substantial manual input for image re-judgment. Re-judgment personnel take an average of 2-3 seconds per image, with high costs, low efficiency, and susceptibility to fatigue and experience differences leading to quality fluctuations.
2) Breaking the limits of the human eye: The panel industry has over 120 types of defects, requiring hundreds of inspectors to judge millions of images daily. Limited by inherent bottlenecks such as eye fatigue, experience differences, and inconsistent standards, the human eye's limitations result in only 85% accuracy. In semiconductor wafer inspection, the system can identify 0.2-20μm macro defects, far exceeding the detection limits of traditional manual or rule-based algorithms.
3) Reducing quality inspection costs: The trial-and-error costs in the pan-semiconductor industry involve equipment processes and bidirectional measurement of techniques, with single defect omission losses potentially reaching tens of millions.
For pan-semiconductor enterprises, quality inspection remains a critical checkpoint related to efficiency and reliability. With the rapid evolution of AI technology, ADC has become the core engine for improving quality and efficiency in manufacturing. GeChuang DongZhi's self-developed Tianshu AI, with its integrated 'Sensing-Storage-Computing-Knowledge' technical architecture, serves as the 'sensory' weapon in this transformation. Through the collaboration of the 'Training-Inference-Scheduling' three modules, it achieves full-process reconstruction of industrial quality inspection.
- Training platform: Built-in tens of millions of defect feature libraries, supporting low-code model iteration.
- Inference platform: Equipped with lightweight edge computing boxes, single image processing speed reaches millisecond level, 5-10 times faster than the human eye.
- Resource scheduling: Dynamically allocates GPU computing power, achieving concurrent operation of 100+ algorithm models at TCL Huaxing.
Based on deep insights into the pan-semiconductor industry's needs for defect detection accuracy, speed, and consistency, GeChuang DongZhi addresses common pain points such as low inspection efficiency, high labor costs, and difficult data tracing through multi-module collaborative architecture design, building an industry-leading optical laboratory. The equipment covers visual inspection imaging needs for 30+ industries. Breakthroughs have been achieved in optical solutions for imaging materials with transmittance, reflectivity, and absorbance, with multiple optical imaging patents applied for, greatly reducing optical imaging technology and algorithm detection indicators. This accelerates customers' experience of triple leaps in quality inspection: 'millisecond-level re-judgment efficiency, over 95% detection accuracy, and cost reduction through model self-training and data self-governance', outperforming competitors with comprehensive advantages.
- Faster production cycle: No reference photos needed, capacity doubled, production more efficient.
- Strong system compatibility: Supports mixed multiple image types, meets diverse inspection needs, offers greater choice freedom.
- Fast, accurate, and stable detection results: Quickly uploads classified results to yield systems, enabling timely access to inspection results, quick decision-making, and improved production efficiency.
- More flexible services: Customizes models based on pan-semiconductor enterprise needs, aligning with unique production processes and yield requirements.
- Faster response: Quickly responds to customers anytime, formulates improvement plans, ensures timely problem resolution.
- Lower expansion costs: Software can be sold separately, reducing enterprise expansion costs, allowing flexible business scaling.
- Strong price competitiveness: Obvious advantages in price and maintenance fees, providing cost-effective solutions, lowering operational costs.
This technological maturity is not accidental. It stems from years of continuous investment and interdisciplinary integration in industrial vision, optical design, precision machinery, and control algorithms. Currently, Tianshu AI has been successfully applied in 1000+ industrial scenarios, validated by the landing of the Excellent Level Intelligent Manufacturing Demonstration Factory, with accumulated billions of defect data providing strong support for model and data self-training and self-optimization.
![](\"https://www.getech.cn/files/edit/20251009/AYzAU8Ob658kyq4pQaN297pZANIM4PS5.png\")
Notably, the company has numerous typical cases in the semiconductor, semiconductor display, and SMT industries. In a semiconductor production project, it helped a client reduce labor input by 90%, with defect detection rates exceeding 99%.
![](\"https://www.getech.cn/files/edit/20251009/Uo5fSX9rjaDTtVr5HkAAY84CXj2mxhZ3.jpg\")
The success of TCL Huaxing (Suzhou)'s AI digital employees is the best footnote to GeChuang DongZhi's industrial quality inspection practices. Currently, the company is actively promoting industry-academia-research collaboration, partnering with TCL Industrial Research Institute to strengthen capabilities in complex defect pattern discovery and cross-domain knowledge transfer.
In the future, as chip sizes continue to shrink and new display technologies accelerate iteration, industrial inspection will evolve towards higher precision, faster response, and stronger generalization. GeChuang DongZhi will serve as a key cornerstone in building China's industrial intelligence system, redefining the role of 'quality inspection' in manufacturing—from passive to preventive, from experience to intelligence-driven—helping Chinese pan-semiconductor enterprises win in global competition with 'intelligence'.
Efficient collaboration between Stocker and AMHS relies on hardware docking, software interaction, and algorithmic decision-making, which break down 'equipment silos' to achieve full-link efficient collaboration and build an intelligent logistics network.
Stable physical connections are the premise of full-link collaboration. Stocker ensures 'seamless docking and zero-interruption collaboration' with the AMHS system through standardized interfaces, multi-carrier compatibility, and full-status sensing.
![](\"https://www.getech.cn/files/edit/20251009/3vbiJCE72wAfrukQV862sXWUTN6uCZVy.jpg\")
Interface standardization: Stocker can be equipped with various interfaces like OHT Port, Manual Port, AGV Port, and CV Port, all using SEMI-standard mechanical structures and electrical protocols to avoid collaboration interruptions due to incompatibility;
Key interface redundancy: Critical interfaces like OHT Port and AGV Port can switch between Input/Output functions; a single port failure does not affect overall logistics, ensuring continuous system operation.
Intelligent carrier recognition: Stocker integrates visual recognition (OCR/BCR) and RFID technology to automatically identify different carriers like FOUP (wafer boxes), Trays, and Cassettes;
Customized grasping: Robotic arms can be equipped with customized grippers, designed based on carrier type and specifications to meet full-process material handling needs from front-end manufacturing to back-end packaging and testing.
Distributed sensor network: Stocker can be configured with sensors for temperature, humidity, pressure, etc., based on storage environment requirements, to collect internal environmental data in real time;
Predictive maintenance: Intelligent analysis based on equipment operating parameters and other data provides early warnings of mechanical failures, avoiding unplanned downtime.
After hardware connection, unified 'communication language' is needed for data interaction. The SECS/GEM protocol, as the 'common language' of semiconductor equipment, supports deep collaboration between Stocker and systems like MCS, RTD, and MES.
![](\"https://www.getech.cn/files/edit/20251009/jST916fX2lj6lVcZXdL4VZnPA1MyW4FF.jpg\")
Information synchronization: Stocker reports its inventory map, storage location status, equipment status, etc., to MCS in real time, allowing MCS to build a global material perspective and know 'which materials are where';
Task execution: MCS issues 'store/retrieve' tasks to Stocker based on MES work order instructions; Stocker feeds back results to MCS after execution;
Congestion avoidance: When MCS detects multiple OHTs queuing at a Stocker entrance, it can dynamically assign tasks to other idle Stockers for load balancing.
Intelligent predictive scheduling: RTD predicts the next batch of material demand time based on equipment production rhythm and instructs MCS to schedule Stocker in advance;
Dynamic path optimization: Real-time analysis of the entire factory's logistics status dynamically adjusts material access priorities and path planning, achieving an advanced collaboration mode of 'materials waiting for machines'.
Work order and material binding: When MES issues a work order, it simultaneously transmits information like Lot number and process parameters to Stocker; Stocker binds the work order to the carrier via RFID, ensuring 'each storage location corresponds to a unique work order' to avoid mixing risks;
Yield data feedback: Stocker uploads material storage duration and environmental parameters (e.g., oxygen concentration, temperature, humidity) to MES; if a batch of wafers has abnormal storage conditions, MES can trace the batch's yield data to assist in analyzing yield loss causes.
True efficient collaboration is not just 'responding to instructions' but 'proactively optimizing'. Stocker enhances the global efficiency of the AMHS system through inventory optimization algorithms, digital twin simulation, and other technologies.
![](\"https://www.getech.cn/files/edit/20251009/81Y8G5Mh65nhcuN1qVxzj3ECMZl06fH4.jpg\")
Adaptive storage location allocation: Dynamically calculates the optimal storage location based on multiple factors like material priority, destination, and equipment status;
Proximity + priority dual strategy: Assigns materials about to be shipped out to the 'fast shipping area' (near the OHT port), high-priority work order materials to the 'priority storage area', and low-frequency access materials to the 'deep storage area', improving space utilization and shipping efficiency.
Virtual debugging: Simulates Stocker operation in a virtual environment through digital twin technology to verify collaboration logic in advance;
Stress testing: Simulates collaboration scenarios during peak periods to verify Stocker interface capacity and MCS scheduling capability, identifying bottlenecks early and optimizing track layout and task allocation strategies;
Forward-looking planning: Adjusts system configuration based on simulation results to avoid congestion and bottlenecks in actual operation.
In the AMHS collaboration field, domestic Stocker companies break international monopolies with 'customization, high cost-effectiveness, and rapid response', with Getech's practices being highly representative.
Getech provides an 'AI+CIM+AMHS' integrated solution, with Stocker deeply collaborating with self-developed OHT, MCS, and MES, overcoming multi-vendor system compatibility issues to achieve end-to-end seamless docking.
Getech Stocker uses self-developed STKC control system, supporting synchronous control of overhead and ground tracks (handling accuracy ±0.5mm), Auto Teaching (automatically acquiring storage location data), with 100% self-developed software architecture, allowing quick customization of collaboration logic based on customer AMHS needs.
Getech Stocker completes project delivery in as fast as 3 months, with 7×24 service response (2-hour response, 4-hour on-site). In a wafer fab AMHS upgrade project, Getech set an industry record with '15 days to deliver the first device'.
From precise hardware docking, to software data interoperability, and algorithmic intelligent optimization, Stocker efficiently collaborates with every node in the AMHS network, performing a symphony of efficient, reliable, and intelligent logistics automation. In the future, with technologies like real-time digital twin simulation of collaboration scenarios and AI large models optimizing global scheduling, Stocker will further push the boundaries of collaboration efficiency, driving AMHS from 'automation' to 'intelligence'.
![](\"https://www.getech.cn/files/edit/20250925/J8TDCIyAVkYSKSWFrYY0j894ZcGb9Bgz.jpg\")
On September 24, the Shanghai National Convention Center was bustling with crowds, and the 25th China International Industrial Fair was in full swing. At this global industrial event with the theme 'Industrial New Quality, Smart Manufacturing Without Borders', Getech's Marketing Director Yang Li was invited to attend the 'AI New Smart Manufacturing, Technology New Era' forum on AI empowering smart manufacturing technology development, sharing the stage with executives from companies like Omron, Advantech, and Inovance Technology to discuss ways to break through in industrial intelligence implementation.
![](\"https://www.getech.cn/files/edit/20250925/DcUHqJggBzXch3JWIDjqoIS1oxHoNus3.jpg\")
Yang Li pointed out that industrial AI implementation needs to overcome multiple challenges such as data, models, cost, and engineering. To address this, Getech proposed the 'ABCDE' industrial AI implementation framework, covering Algorithm, Big Data, Computing, Domain knowledge, and Equipment across five dimensions, systematically advancing the closed-loop implementation of AI technology in complex industrial scenarios. Yang Li emphasized that only enterprises with domain knowledge and equipment experience in specific industrial sectors can integrate data, algorithms, and computing power to ground industrial intelligence, meeting the extreme demands of advanced manufacturing in stability and accuracy.
![](\"https://www.getech.cn/files/edit/20250925/EHnIUYe34eijkgQ23X3B1iS7G9Ko4H2O.png\")
Getech has always focused on the semiconductor industry as a strategic area, and based on the 'AI+CIM+AMHS' integrated software-hardware industrial intelligence solution, it helps the semiconductor industry achieve production goals of 'high rhythm, zero defects'. At the soft power level of AI empowerment, Yang Li revealed the evolution path of industrial AI-driven semiconductor CIM, indicating that semiconductor CIM is transitioning from 'experience-driven' to 'data-driven + AI collaboration', relying on the deep integration of large model agents with enterprise-level data, business processes, and system tools to accelerate the arrival of next-generation CIM.
![](\"https://www.getech.cn/files/edit/20250925/81EnyTM5yXythGUV4ghY0hE39JZCthjK.png\")
At the hard power level, Getech is accelerating the deep integration of industrial AI with hardware equipment. In core scenarios of semiconductor material handling, Getech's independently controllable AMHS solutions, represented by OHT (Overhead Hoist Transport) and Stocker, have secured several major Chinese semiconductor clients. Based on the first-phase project of the Hong Kong University-Getech Industrial Joint Laboratory, they co-developed intelligent scheduling models that perceive and integrate MES, MCS business data, and OHTC operation data, forming highly business-aware AMHS solutions, aiming to break through in the core area of thousand-unit OHT scheduling algorithms. Clearly, the integrated software-hardware capability is Getech's core advantage distinguishing it from pure software vendors.
![](\"https://www.getech.cn/files/edit/20250925/OUhUhDvBAMuKTWRapodS7NOEc3TKgU6P.png\")
Additionally, as an industrial AI leader strategically incubated by TCL, Getech focuses on the 'AI+Industrial Software+Smart Equipment' strategic framework, accumulating rich typical cases across multiple scenarios in TCL's various industries. Yang Li shared that in serving TCL CSOT, a leader in semiconductor display, the company achieved deep industrial AI empowerment based on smart application scenarios across five digital platforms: R&D, supply chain, product, sales, and ERP. In quality management scenarios, Getech uses AI industrial vision inspection systems to automatically detect, classify, and handle defects, increasing defect detection accuracy to 95%. In product R&D scenarios, Getech collaborates with TCL industries, using knowledge graphs and vector retrieval technology to integrate massive process documents, equipment parameters, and anomaly records into a 'virtual expert system', addressing core needs such as intelligent Q&A, new employee training, product issue analysis, and document assistance in the display field, helping engineers provide reliable knowledge Q&A and decision support, with document summarization efficiency improved by 60% and product issue analysis efficiency by 20%.
Notably, Getech recently launched the CIM AI Foundation multi-agent system platform, providing comprehensive industrial agents and large model tools for CIM systems, lowering the barrier to integrating large model technology into CIM systems, and accelerating cost reduction and efficiency improvement in semiconductor manufacturing. In practical applications, the platform's GT Insights conversational ChatBI function saves 90% of data query, analysis, and interpretation time; the AI-enhanced group FDC solution helped a leading semiconductor enterprise achieve cross-factory equipment difference analysis and specification optimization.
![](\"https://www.getech.cn/files/edit/20250925/XZ4FBKB8fN6TIvqNdkvu28vfxgeeU4bo.jpg\")
Getech's voice at the CIIF represents a company's technological confidence. Its deep empowerment of TCL's internal industries has become a microcosm of Chinese manufacturing moving from 'scale' to 'quality'. It not only provides the industry with a reusable methodology for industrial AI implementation but also uses the CIIF as a window to showcase to the world China's path for industrial intelligence moving from 'concept first' to 'value practice' in a new stage.
As a leading enterprise in industrial intelligence, Getech continues to chart a benchmark path for China's industrial intelligence development through its integrated 'AI+Industrial Software+Smart Equipment' capability, helping Chinese manufacturing establish a voice in the global smart manufacturing competition.
On September 16, TCL Technology won the Fifth China Quality Award with its 'extreme, leading, collaborative' quality management model, becoming the only award-winning enterprise in Guangdong Province and the Greater Bay Area this session. This honor not only fully recognizes TCL's quality management system, but also reflects the deep empowerment and value of Getech, the core force behind intelligent manufacturing and digital transformation, in industrial quality digital intelligence.
![](\"https://www.getech.cn/files/edit/20250925/PekUBUwH9DIPUlPyIEiqC5JSrdYjDxoJ.jpg\")
As a leading industrial intelligence enterprise strategically incubated by TCL, Getech has long surpassed traditional 'software service providers' to become an AI-driven industrial intelligence solution provider deeply integrated with AI, industrial software, smart equipment, and industry know-how. Over the past seven years of accompanying TCL's digital transformation, it has accumulated rich quality management scenario cases.
In TCL CSOT's intelligent manufacturing system, Getech achieved intelligent breakthroughs in quality inspection through the ADC automatic defect classification system and AOI automatic optical inspection equipment. Facing the pain points of numerous defect types, low judgment efficiency, and high reliance on manpower in the panel industry, Getech's self-developed Dubhe-S200H intelligent visual inspection system, integrating deep learning and traditional algorithms, accurately captures subtle defects such as dots, lines, and Mura on screens, with a detection rate exceeding 95% and accuracy above 99%. The system seamlessly integrates with MES and SPC systems, enabling real-time data upload, analysis feedback, and closed-loop management, not only significantly improving inspection efficiency and consistency but also bringing tens of millions in annual benefits to TCL CSOT.
Additionally, the YMS yield management platform co-developed by Getech and TCL CSOT has been successfully applied to 3 major production bases and 9 factories. By integrating yield-related defect data, equipment parameter data, and production history data into the platform, 111 functions under modules such as yield monitoring, defect analysis, and process analysis were developed, supporting yield management engineers in quickly locating yield anomalies and conducting multi-dimensional root cause analysis efficiently. Among them, YMS innovatively applies CTQ analysis methods, accumulating yield analysis case experience, laying a foundation for full-chain automatic monitoring and analysis in factories. Meanwhile, YMS integrates multi-factor analysis tools, big data analysis tools, and generative AI, breaking through traditional analysis methods to further enhance the possibility of identifying abnormal root causes in yield analysis. Compared to the old EDA system, YMS improves data query and analysis performance by over 3 times; reduces yield anomaly analysis time from 4 hours to 0.5 hours, improving efficiency by over 80%, with overall platform benefits reaching 13 million yuan.
In Zhonghuan Leading's semiconductor material production环节, Getech built a full-process quality traceability system through the QMS quality management system platform. It addressed the informatization gap in 'IQC management', optimized the 'issue management' module, accelerated 'change management' construction, and improved quality management informatization levels. This promoted data sharing and collaboration efficiency by about 15%, optimized the issue management module to reduce quality losses by about 100,000 yuan annually, and won long-term trust from semiconductor customers like Ruijiewei through quality stability during industry downturns.
In TCL Zhonghuan's photovoltaic module manufacturing sector, Getech developed a management cockpit for quality data collaboration, traceability, and unification based on its mature self-developed QMS system, improving quality management efficiency by 20%. Notably, the 'Pathfinder' business analysis system created by Getech for TCL Zhonghuan's photovoltaic business line extends from quality management to operational quality. By integrating data from MES, QMS, ERP, and other systems, it enables real-time monitoring and cross-factory benchmarking of key quality indicators such as breakage rate, conversion efficiency, and degradation rate. Managers can drill down from the group level to production lines, shifts, or even equipment levels, quickly identifying quality fluctuations and driving improvements. This system not only helped TCL Zhonghuan achieve refined management breakthroughs during industry slumps but also became a core engine promoting its transition from 'factory-level quality management' to 'group-level quality collaboration'.
In TCL's smart terminal领域, Getech's ADC, QMS, and other systems act as 'AI judgment digital employees' and 'AI quality managers' in the production processes of TCL air conditioners and TCL home appliances.
Getech's empowerment in quality management is reflected not only at the single technology or system level but also in deep integration and ecosystem building across the entire manufacturing process. From defect detection in TCL CSOT's panels to quality management in Zhonghuan Leading's silicon wafers, and cross-base quality benchmarking in TCL Zhonghuan's photovoltaic industry, Getech uses integrated software-hardware solutions like ADC, QMS, SPC, YMS, LIMS, and AOI to shift quality management from 'post-inspection' to 'real-time prevention' and from 'experience-driven' to 'data-driven', building a quality moat for TCL's manufacturing system.
TCL's winning of the China Quality Award is a significant milestone in the transformation from 'Made in China' to 'Quality Made in China', and the best recognition of Getech's years of dedication to industrial quality digital intelligence. In the future, Getech will continue to innovate in visual inspection, yield optimization, quality traceability, and other areas, helping TCL and even Chinese manufacturing enhance quality discourse power in global competition.
From September 23 to 27, the 25th China International Industrial Fair, one of the world's largest, most comprehensive, highest-level, and most influential industrial events, will grandly open in Shanghai Hongqiao. With the theme 'Industrial New Quality, Smart Manufacturing Without Borders,' it focuses on the main lines of 'high-end, intelligent, and green,' starting a grand narrative of global industrial intelligence. The exhibition, covering 300,000 square meters, has attracted 3,000 exhibitors from 28 countries and regions, including nearly 300 Fortune 500 and industry leaders like Siemens, Schneider Electric, Bosch, ABB, and over 700 specialized, refined, and innovative enterprises. Visitors from 102 countries and regions will showcase the demand and charm of 'Made in China.'
Getech's Marketing Director Yang Li is invited to attend the AI-enabled Smart Manufacturing Technology Development Forum themed 'AI New Smart Manufacturing, Technology New Era' during the Industrial Fair. Together with executives from companies like Omron, Advantech, and Inovance, they will decode strategies for integrating AI into smart manufacturing. Let's look forward to contributing 'Eastern Wisdom' and 'Chinese Solutions' to the global industrial transformation and upgrading.
![](\"https://www.getech.cn/files/edit/20250923/IRY6U791MDeW8lMfUJfXMs5rlrBx1eYG.jpg\")