The company's business continues to grow rapidly. In 2022, the company achieved revenue of 1,558 million yuan, an increase of 80.9% over the previous year, and the revenue CAGR from 2019 to 2022 reached 89%.
The company's gross margin was 32.6%, up 1.6 percentage points from 2021; the net loss of the parent was 360 million, while the adjusted net loss was 138 million (adjustments were share payment costs, listing expenses, amortization of intangible assets and changes in financial assets arising from acquisitions), and the adjusted net loss ratio was 8.9%, down 7.6 percentage points from 2021. We believe that strong demand for applications is fundamental to the implementation of AI. The performance that has maintained rapid growth over the years is a market validation of the company's strong demand in the AI field.
A typical example of the industry of large industrial model+small sample learning. In 2022, the company's “AI+ Manufacturing” business segment revenue was 948 million yuan, an increase of 111% over the previous year, accounting for more than 60% of total revenue. The company has achieved remarkable results in key manufacturing fields such as automotive equipment, 3C high-tech, panel semiconductors, steel and metallurgy, energy and electricity, engineering construction, food and beverage, and new materials. Through the implementation strategy of 1xN replication and 1+N expansion, as well as the “point-line-plane” business development idea, it has formed a sustainable rapid growth path. According to IDC data, the company ranked in the leaders quadrant of China's industrial data intelligence market in 2022, ranked the 2nd largest AI industrial quality inspection vendor in China for three consecutive years, and was the only company that ranked in the top four in both computer vision and machine learning fields.
Manufacturing AIGC can be expected. The company recently released AinOGC products, which are AIGC products for the manufacturing industry. AIGC is currently a hot word, especially in the fields of large multi-modal models such as text generation, image generation, and code generation. However, with regard to the TOB scenario, especially in the complex closed-loop application field of manufacturing, how to implement the scenario, how to further improve the intelligent efficiency of AI, and actually reduce costs and increase efficiency for enterprises is the key to strong demand for AIGC in the manufacturing industry. Based on practical accumulation on the MMOC platform, including technical assets such as data sets, algorithm libraries, model libraries, etc., the company is gradually developing large industrial pre-training models and AIGC products for the manufacturing industry, that is, large industrial models+AInnoGC engines. Recent implementation scenarios: (1) BI (Management Intelligence Analysis) and MI (Management Intelligence Analysis) Q&A and automatic generation of statistical reports. Through the NLP dialogue mode, various questions in the production process are presented to AinNogc. AinOGC will automatically generate answers and statistical charts based on the questions, such as: production line efficiency daily/weekly reports, production line operation plans, analysis and records of the causes of equipment downtime, analysis reports on factors affecting production efficiency, energy saving planning and statistics, measures to achieve energy saving goals, etc.
(2) Intelligent manufacturing training center. Through the AinNoGC platform, the training instructor automatically generates tens of thousands of defect sample data for students to practice the training model, and rates the students through AinNoGC. AinNoGC will give knowledge points for intensive training; through interaction with AinNogc, the participants will obtain system programming knowledge points (for example, using C++ to write a node in the ROS system to control the movement of the robot arm), control program debugging, etc.; in addition, on the training platform, it is also possible to automatically generate virtual industrial scenarios through AinNoGC courseware, etc.
Medium- to long-term implementation scenarios: (1) Automatic virtual simulation generation of industrial production line design and product process insight. The company can automatically generate production line designs based on existing digital twin technology solutions by combining AinNogc, a large-scale industrial model. Customers can describe requirements and pain points through the NLP dialogue interface, and AinNogc will automatically generate simulated production line design plans at the process level and production line design process level. The entire process will involve a large number of industry challenges.
(2) Automatic deployment of flexible manufacturing of industrial robots. For example, AGVs in the field of silicon wafer manufacturing can interactively enter instructions through AinNOGC and flexible robots. Flexible robots can automatically generate algorithms and models (self-programming/adaptive) and automatically launch online work, eliminating a large amount of manual coordination and deployment work.
The company began moving from AI1.0, a delivery solution with MMOC as a platform, to AI2.0, a generative solution centered on the big industrial model+MMOC+aInnoGC.
Investment advice. In 2023, the average PS of comparable companies was 15 times, the average PS of American companies was 10 times, and Shangtang's PS was 13 times. Innovation Qizhi's 21-24 CAGR growth rate was higher than other AI companies, but its gross margin was lower than that of other comparable companies.
We forecast that the company's revenue for 2023-2025 will be 22.71 (-8.98%) /33.37 (-16.49%) /4,895 billion yuan respectively, and the revenue CAGR for the 2021-2024 period will be 57% (-14.93%). Taking into account factors such as the company's and comparable companies' revenue growth rate, gross margin level, industry position, and the company's differentiated competitive advantage in the industry and track, we gave Innovation Qizhi 8 times PS in 2023 (originally 11.5x in 2022). The target market value was RMB 18.167 billion (+1.84%). Based on the total share capital of 559.3 million yuan, the target price was RMB 32.48 per share (corresponding to HK$37.12 per share, RMB 1.82% increase, HKD 2.23% increase, difference between RMB and HKD 2.23%. (Caused by exchange rate changes), maintaining a “superior to the market” rating.
Risk warning. The slowdown in the company's revenue in the manufacturing sector affected the overall revenue growth rate, increased competition in the industrial quality inspection sector caused the company's revenue growth rate and gross margin to fall short of expectations, and factors such as the impact of overseas restriction policies on relevant market valuation levels