Industry opens the era of intelligence, and Simou Technology has become a new unicorn in intelligent manufacturing

Industry opens the era of intelligence, and Simou Technology has become a new unicorn in intelligent manufacturing

#Industry #opens #era #intelligence #Simou #Technology #unicorn #intelligent #manufacturing

“Water is big and fish is big”, a unicorn reappears in the field of intelligent manufacturing.

In June 2021, AI technology company Simou Technology completed a US$200 million Series B financing and announced that it has become a unicorn company. As a new force that promotes the intelligent transformation of the industry, Simou Technology has also become another new unicorn in the field of intelligent manufacturing following the big platforms such as Kaos and Shugen Internet.

According to the National Bureau of Statistics, in 2019 and 2020, the scale of my country’s industrial added value will exceed 31 trillion. An efficiency improvement of 1-2% in the industrial field means hundreds of billions of economic value-added. At the same time, many manufacturing enterprises are facing the current situation of labor shortage and recruitment difficulties, and need to upgrade technology to alleviate production and management problems. Therefore, cutting-edge technologies such as AI are increasingly being applied to the improvement of industrial efficiency. With the promotion of new technologies and the demands of internal industrial upgrading, a new round of industrial revolution is brewing.

According to Jia Jiaya, founder and chairman of Simou Technology, intelligence is the future of industrial manufacturing. Based on the development and application of AI full-chain technology, Simou Technology will also actively participate in the transformation of the entire production method.

Jia Jiaya, Founder and Chairman of Simou Technology

Received 200 million US dollars in financing, Simou Technology has become a new unicorn

Simou Technology was established in December 2019. Since its establishment, it has completed four rounds of financing. The recently completed US$200 million Series B financing has been continuously increased by old shareholders such as IDG, Cornerstone, Sequoia China, Songhe, Lenovo Ventures, and Zhen Fund. It has also attracted new shareholders with industrial resources and Xuan Capital, The participation of Niu Capital and Shenwan Capital. It is understood that this round of financing will be mainly used to further deepen the market, increase investment in the research and development of intelligent manufacturing products, and promote the large-scale implementation of more scenarios.

At the beginning of its establishment, Simou Technology has received a lot of attention from the industry.

Jia Jiaya, the founder and chairman of Simou, is a tenured professor at the Chinese University of Hong Kong, an academician of the Institute of Electrical and Electronics Engineers (IEEE), an associate editor of the industry’s top journal TPAMI and an editorial board member of IJCV. As an authoritative expert in the fields of computer vision, artificial intelligence, computer graphics, and computer imaging, Jia Jiaya has published more than two hundred top-level papers during his more than 20 years of research career, and his papers have been cited more than 30,000 times. Jia Jiaya joined Tencent in 2017 and is the highest-level outstanding scientist of Tencent. He was the founding head of X-Lab of Tencent Youtu Lab, providing artificial intelligence technology for Tencent platform. The AI ​​algorithm product interface he developed still maintains to this day. High-level call volume.

In addition, the company’s co-founder and CEO Shen Xiaoyong is Jia Jiaya’s PhD student in Chinese in Hong Kong. The technical team of Simou Technology has also been deeply involved in the field of computer vision for many years. The core members are all from the world’s TOP30 universities, well-known industrial Internet companies and first-class industrial automation companies.

Jia Jiaya once publicly stated that the original intention of creating Simou Technology was because “the sinking of AI will definitely bring new flashpoints.” After many investigations and in-depth thinking, Simou chose to apply AI technology to the two fields of industry and ultra-high-definition video first. In the construction of its technical system, Simou Technology emphasized the construction from a single algorithm to the system architecture, and launched a product matrix integrating software and hardware to provide the industry with end-to-end visual AI full-stack services.

In the industrial field, Simou has launched two platform products, the SMore ViMo industrial AI platform and the SMore ViLab toolbox. Based on the common characteristics of customers in various downstream industries, it has developed smart boxes, smart cameras, barcode scanners and other suitable devices. Standardized smart hardware. At present, more than 30 intelligent manufacturing software and hardware integrated products and corresponding solutions developed by it have been used in semiconductor, precision optics, new energy, automobile manufacturing and other industries, and have served Carl Zeiss, Airbus, Bosch, Canon Waiting for more than 100 top customers.

Simou Technology AI software and hardware integration products

At this stage, the intelligent manufacturing business line accounts for 70% of the company’s revenue, and the other 30% of the revenue mainly comes from ultra-high-definition video. With a global business base, Simou Technology also has R&D and business centers around the world such as Hong Kong, Shenzhen, Shanghai, Suzhou, Beijing, Hangzhou, Chongqing and Tokyo, Japan.

AI starts the era of industrial intelligence, and Simou Technology has served hundreds of top customers

Historically, the development of manufacturing is inseparable from the promotion of new technologies at various stages. After mechanical automation in 1940 and electronic informatization in the 1970s and 1980s, the manufacturing industry gradually advanced from manual operations to preliminary industrial control automation. However, the demand for liberating productivity and improving efficiency still exists for a long time, and new technologies represented by AI have begun to promote the development of manufacturing in the direction of intelligence.

In the view of Jia Jiaya, founder and chairman of Simou Technology, “industrial transformation comes from the needs of the times.” For manufacturing companies, the first thing that AI can achieve is to replace labor and use less labor to achieve the same or even higher work efficiency. At the same time, in the past, relying on human experience to complete process upgrades was time-consuming and uncertain, while AI can conduct large-scale automatic testing and prediction of materials and process flows, and quickly accumulate experience through deep learning.

The landing of AI in the manufacturing industry is inseparable from the combination of specific scenarios, and Simou has chosen the semiconductor industry as the first stop for technology landing. Compared with other industrial categories, the semiconductor industry has a higher degree of automation and a good data foundation, but because its manufacturing process is extremely complicated and long, and the yield of each process step must be close to 100%, in order to ensure that the final production yield is maintained at a constant level. Acceptable level. Jia Jiaya described the difficulty of defect detection for semiconductor products as “in the entire Guangdong Province, a traffic light with a slight fault is located within one second.” Therefore, how to improve the efficiency of quality inspection has always been a major problem in semiconductor manufacturing. Due to the high technical threshold, most manufacturers are also discouraged.

Relying on the core team’s many years of accumulation in the AI ​​field and the momentum of a newborn calf, Simou took over the two hard bones of the industry that the industry dare not touch easily at the beginning of its establishment. The results were unexpected. In just two weeks, Simou completed data collection and analysis, model training, software development and online, and delivered defect detection and process analysis products within one month. The effect exceeded customer expectations by 20%. Facts have proved that Simou’s AI technology capabilities can indeed solve the problems that have plagued the manufacturing industry for many years, which also makes Simou’s road to AI + intelligent manufacturing industrialization more and more smooth.

After that,Simou Technology has successively entered various fields such as precision optics, automobile manufacturing, new energy, and consumer electronics.In the process, Simou discovered that different downstream customers have different needs, but there are still many commonalities that can be refined in the same type of scenario. For example, in the field of precision optics, the production processes of automotive glass, spectacle lenses, and camera lenses are very different, and the underlying technical logic is similar. Based on the same technical logic, Simou can provide universal products and solutions for the same type of enterprise.

The core algorithm capabilities of Simou Technology have realized defective pixel-level recognition, text detection and recognition, target analysis/location technology, and multi-category distinction functions. Based on the algorithm, Simou launched an AI solution for the entire manufacturing process from the supply of materials to the warehouse, incoming material inspection, production line production, quality inspection to outbound delivery. Through the common refinement of downstream industries, Simou has developed more than 30 software and hardware integrated products.

Source: Simou Technology

Hardware products are mainly used for data collection and analysis on the edge, including smart code scanning devices, smart cameras, edge AI boxes, and smart handheld devices. By bringing AI algorithms such as deep learning to edge devices, Simou complements an important part of edge-cloud collaboration, and its simplified system logic also makes management and traceability clearer.

The software algorithm itself is the advantage of Simou, and the transformation of AI from a single algorithm to the direction of the architecture is also the direction of technological research and development of Simou. To put it simply, using a single point of technology as the direction of force can easily lead to homogeneity and duplication of products and technologies. By constructing an AI system architecture and integrating algorithms into different modules in a systematic mode, it is easier to summarize commonalities and experiences, thereby improving efficiency. For manufacturing industries with segmented scenarios and diverse needs, a systematic software platform is particularly critical.

Simou Technology has independently developed two platform-based software, namely SMore ViMo Industrial AI Platform and SMore ViLab Toolbox. The SMore ViMo industrial AI platform is a full-process industrial vision AI end-to-end system consisting of two systems: model training and model operation. The SMore ViLab toolbox is a software and hardware integrated toolbox for the rapid delivery and operation of visual AI projects based on deep learning algorithms. It can quickly perform Demo demonstrations and POC delivery. The platform is compatible with multi-manufacturer, multi-interface type cameras and other hardware. Users can quickly and easily connect to external lighting imaging systems, flexibly deploy deep learning algorithms and run tasks.

Jia Jiaya concluded that Simou’s software products are similar to the Windows operating system and can automatically adapt to the apps required by the factory. The hardware is like a PC, mainly used for data collection and interaction. The integration of software and hardware opens up the entire chain of data collection, analysis, and algorithm precipitation.

With the blessing of platform-based software and hardware products, the delivery time of Simou Technology at the customer’s premises has also been shortened from 3 months to half a year last year to about 1 month. At this stage, more than 30 smart manufacturing software and hardware integrated products of Simou have been implemented in production scenarios including airplanes, automobiles, new energy batteries, smart phones, smart wearables, chips, precision optics, and new-generation display technologies. Hundreds of top customers.

In the semiconductor field, Simo has cooperated with a number of national key enterprises to launch dozens of AI+ solutions and integrated software and hardware equipment from wafer inspection to PCB inspection to chip process analysis. Especially in the analysis, identification, and judgment of defect types, the current accuracy rate of Simou has reached 99.99%, and it has realized the closed-loop control from equipment process optimization to product quality improvement.

In the field of automobile production, Simo put its bearing AI inspection integrated machine into a test run on the production line of a Fortune 500 auto parts manufacturer in October last year. The launch of this product can automatically identify 23 types of defects at one time, helping customers to improve the efficiency of quality inspections. It has increased by more than 80%, and the detection accuracy is close to 100%.

From smart manufacturing to ultra-high-definition display, creating the end of the industry for “unmanned factories”

From nano-scale semiconductors to cars and airplanes measuring tens of meters or hundreds of meters, Simou Technology is crossing physical boundaries and applying AI technology to all aspects of manufacturing.

Jia Jiaya believes that intelligent manufacturing is an intelligent process, with the focus on “intelligence.” Platform products based on AI architecture provide strong internal system support for intelligent manufacturing products and services. However, Simo can break through multiple industries in a short period of time. On the one hand, it is inseparable from the way of thinking of the same frequency as the manufacturing factory, that is, from the perspective of the factory, starting from the practical problems such as the difficulty in recruiting workers, the possibility is reversed AI products and solutions. In addition, a deep understanding of know-how in various industries is also essential.

Many industrial Internet companies emphasize data collection and analysis, but first clarify “what is data”, that is, how to find key data based on a deep understanding of downstream enterprise customers.

In the process of serving customers, Simou has formed an automatic adaptation platform capability for more than 1,000 hardware devices, and can independently develop software and hardware integrated data collection and analysis full-chain solutions for customers in different fields and different scenarios in a short period of time. In order to resonate with the downstream industry, Simou also upgraded from the original software algorithm company to a supplier of integrated software and hardware products and solutions.

In addition, for manufacturing companies, in addition to intelligent production and operation processes, higher-level display systems are also needed. In the process of serving customers, Simou found that customers have a huge demand for front-end display equipment and display screens, but lack of professional HD video team production content. Considering the wide application prospects of ultra-high-definition video in the fields of culture, education, entertainment, intelligent transportation, broadcast and television, Simou Technology has begun to develop ultra-high-definition video services that are also based on AI technology. For manufacturing customers, Simo Technology has completed the closed loop from back-end intelligent manufacturing deployment to front-end display.

It is understood that Simou has launched the “Zhi Zhen” ultra-high-definition video AI solution in the field of ultra-high-definition video, serving the entire chain of video content from production to presentation. In addition to manufacturing, the transportation sector is also the main source of customers in this field.

“The new round of industrial revolution must be intelligence.” Jia Jiaya believes that in the future, intelligence will become a universal concept in the manufacturing industry, and it will be an indispensable part of all walks of life. With the continuous deepening of AI technology in the manufacturing industry, factories will gradually realize the “three modernizations” of intelligence, automation and digitization, and will eventually give birth to the emergence of “smart factories”. Specifically, people only need to master key information as top managers, and the entire production decision-making process is completed by AI. At that time, the production line will also switch from a fixed mode to a modular operation to promote the realization of flexible production.

In the process of industrial intelligence, Simou Technology is positioned as the leader of the industry, and at the same time it does not rule out the transformation from a service provider to a changer. Jia Jiaya said that in the future, Simu Technology hopes to break away from the existing production model of the manufacturing industry and design a complete AI production line from start to finish. Once implemented, the original production rules of the manufacturing industry, such as the logic of capacity ramping, may be subverted, and the speed at which new technologies such as AI will transform traditional industries will also increase significantly. Next, Simou will work with partners from all walks of life to jointly explore more possibilities for intelligent manufacturing.

Represented by computer vision, deep learning, natural language learning, etc., artificial intelligence is generally believed to lead a new generation of technological revolution. According to statistics, the scale of China’s artificial intelligence market reached 185.8 billion in 2020, and it is expected that this number will further rise to 10457 billion in 2025.

Jia Jiaya said that at the beginning of its establishment, Simou’s vision was to become a full-scenario intelligent company that leads technology and serves the world. Next, Simou will not define its own boundaries, but will use intelligent manufacturing and ultra-high-definition video as the entry point to actively deploy the underlying AI system architecture capabilities. In the future, Simou will continue to explore a broader development space for artificial energy.


Related Posts