In the context of the “Industry 4.0” era, the manufacturing sector in China is undergoing a critical transformation. Industrial robots, often regarded as the gem of modern manufacturing, have emerged as pivotal tools in this evolution. As researchers immersed in technological innovation, we aim to delve into the industrial robot industry in China through the lens of patent maps. This approach allows us to extract valuable technological intelligence by analyzing patent data, which serves as a reliable indicator of innovation trends and competitive landscapes. The significance of patent literature cannot be overstated—it reduces R&D costs, enriches technological content, minimizes sunk costs, and provides风向标 for policy-making. Our study leverages the Derwent Patent Database (DII) as the primary data source, utilizing tools like TDA and Matlab for comprehensive data mining and visualization. By constructing a detailed patent map, we seek to uncover insights into China’s industrial robot专利布局, technological cycles, and future hotspots, thereby contributing to strategic planning in this vital sector.
The concept of patent maps dates back to the 1960s, initially developed in Japan and later adopted globally for科技战略 and industry research. A patent map is a graphical representation of patent information that facilitates the analysis of technological trends, competitive landscapes, and innovation gaps. In our work, we build upon the framework proposed by Yoon B U, Yoon C B, and Park Y T, which categorizes patent maps into技术地图,管理地图, and权利地图. However, based on the structure of the Derwent database, we have refined this classification to better suit our analysis of China robot专利. Our improved framework focuses on two main categories: the专利技术地图, which includes basic patent information such as application years, institutions, and technological distributions, and the专利管理地图, which delves deeper into专利技术周期 and hotspot predictions through citation analysis and life cycle modeling. This adjusted approach enables a more targeted investigation into the China robot industry, aligning with the unique features of the Derwent data, such as IPC classifications, Derwent manual codes, and patent family linkages.
To gather relevant patent data, we conducted a search in the Derwent Patent Database, focusing on industrial robot-related patents. The search strategy involved using keywords and logical operators to filter out non-industrial机器人 applications. Given the broad scope of robotics, we excluded patents from non-relevant IPC sections, such as A (human necessities), C (chemistry), and others, while emphasizing IPC codes in B (performing operations; transporting) and G (physics). The search formula was: ts=(robot) not ip=a* not ip=c* not ip=d* not ip=e* not ip=f* not ip=h* not ip=b01* not ip=b08* not ip=b09* not ip=b31* not ip=b60* not ip=b62* not ip=b63* not ip=b64* not ip=b66* not ip=b67* not ip=b41* not ip=b42* not ip=b43* not ip=b44*. The time span covered all patents up to 2015, acknowledging an 18-month lag in data completeness, which means data from 2013 onward may be incomplete but still useful for trend analysis. This retrieval yielded a substantial dataset of patents, which we then processed to examine the China robot landscape in detail.
Our analysis begins with the distribution of patent applications by priority year in China. As shown in the data, the number of patents has grown steadily and rapidly, particularly after 2010. The first industrial robot patent in China was filed in 1984, with only one application. Growth was slow until 1994, after which patenting activity became more consistent, with no years of zero applications. From 2005 onward, the growth accelerated significantly, with applications increasing from 140 in 2006 to 1,951 in 2013. This trend reflects the rising importance of the China robot industry in the global context. To quantify this growth, we can model the cumulative number of patents over time using a logistic function, which is common for technology life cycle analysis. The logistic model is expressed as:
$$ P(t) = \frac{K}{1 + e^{-r(t – t_0)}} $$
where \( P(t) \) is the cumulative number of patents at time \( t \), \( K \) is the saturation level (maximum number of patents), \( r \) is the growth rate, and \( t_0 \) is the inflection point. For China robot patents, our拟合 results indicate a rapid growth phase from 2001 to 2015, with an average annual growth rate of 39%. This aligns with national policies and increased R&D investments in the China robot sector.
Next, we examine the distribution of patent applicants in China. Applicants are categorized into four groups: universities, research institutes, enterprises, and individuals. Our statistics reveal that enterprises account for 54.23% of total applications, followed by universities at 29.56%, individuals at 10.82%, and research institutes at 5.39%. However, when looking at the average number of patents per entity, universities lead with 10.40 patents per institution, compared to 2.91 for enterprises, 4.07 for research institutes, and 2.54 for individuals. This suggests that while enterprises file more patents in aggregate, universities are more active on a per-entity basis, highlighting their role as innovation hubs in the China robot ecosystem. The top 15 applicants, listed in Table 1, account for 18.75% of total patents, with universities and research institutes dominating the list. For instance,清华大学 (Tsinghua University) leads with 155 patents, followed by鸿富锦精密工业 (Hongfujin Precision Industry) with 125 patents. This distribution underscores the need for stronger industry-academia collaboration to commercialize innovations in the China robot field.
| Applicant | Number of Patents |
|---|---|
| Tsinghua University | 155 |
| Hongfujin Precision Industry (Shenzhen) | 125 |
| Shanghai Jiao Tong University | 116 |
| Beijing University of Technology | 112 |
| Guangxi University | 101 |
| Harbin Institute of Technology | 88 |
| Suzhou Industrial Park Vocational College | 84 |
| Hon Hai Precision Industry | 76 |
| Shanghai University | 69 |
| State Grid Corporation of China | 65 |
| Tianjin University | 64 |
| Shenyang Institute of Automation, CAS | 63 |
| South China University of Technology | 63 |
| Yanshan University | 60 |
| Shenyang Siasun Robot & Automation | 56 |
To understand the technological focus of China robot patents, we analyze the International Patent Classification (IPC) codes. The top IPC subclasses, as shown in Table 2, reveal that B25J (manipulators; industrial robots) and B23K (soldering; welding) dominate, accounting for 52.46% and 15.52% of patents, respectively. Other subclasses, such as B65G (conveying) and B23P (metal-working other than machining), have smaller shares but show emerging trends. For example, B23P and B05B (spraying) have seen recent growth, indicating new technological directions in the China robot industry. The concentration in a few IPC subclasses suggests that innovation is clustered around core technologies, but the gradual emergence of other subclasses points to diversification. We can quantify this concentration using the Herfindahl-Hirschman Index (HHI), a measure of market concentration, adapted for technology分布:
$$ HHI = \sum_{i=1}^{N} s_i^2 $$
where \( s_i \) is the share of patents in IPC subclass \( i \), and \( N \) is the total number of subclasses. For China robot patents, the HHI value is high, reflecting a focused technological landscape. However, as new subclasses gain traction, this index may decrease, signaling broader innovation in the China robot sector.
| IPC Subclass | Number of Patents | Percentage |
|---|---|---|
| B25J | 3,628 | 52.46% |
| B23K | 1,074 | 15.52% |
| B65G | 482 | 6.97% |
| B23Q | 345 | 4.99% |
| B62D | 212 | 3.06% |
| G05B | 198 | 2.86% |
| B21D | 165 | 2.39% |
| B23P | 142 | 2.05% |
| B05B | 128 | 1.85% |
| B65B | 115 | 1.66% |
In addition to IPC codes, we examine the application fields using Derwent manual codes. These codes represent the practical applications of patented technologies. The top Derwent code combinations, listed in Table 3, show that P62 (hand tools, cutting) is the most frequent, appearing in 20.93% of patents. Combinations involving P62 with X25 (industrial electric equipment) or T06 (process and machine control) are also common, indicating integrations in cutting tools and electrical systems. Single-code combinations account for 29.4% of the top 15, while multi-code combinations suggest跨领域 innovations. This分布 highlights that China robot technologies are applied in specific domains like cutting and welding, but there is potential for expansion into other areas. To assess the diversity of applications, we can calculate the Shannon entropy index:
$$ H = -\sum_{i=1}^{n} p_i \log_2 p_i $$
where \( p_i \) is the proportion of patents in Derwent code combination \( i \). A higher entropy value would indicate greater diversity in the China robot application fields, which is currently moderate but evolving.
| Derwent Code Combination | Number of Patents | Percentage |
|---|---|---|
| P62 | 1,448 | 20.93% |
| P62, X25 | 396 | 5.72% |
| M23, P55 | 180 | 2.60% |
| P62, T06, X25 | 171 | 2.47% |
| Q35 | 166 | 2.40% |
| P62, T01, T06, X25 | 137 | 1.98% |
| M23, P55, X24 | 134 | 1.94% |
| P56 | 132 | 1.91% |
| P42 | 131 | 1.89% |
| P52 | 87 | 1.26% |
A critical aspect of our study is the analysis of patent technology life cycles and hotspot predictions. Using the logistic model, we forecast the life cycle of industrial robot patents globally and for China specifically. For the world, the life cycle comprises four stages: introduction (1963-1980), growth (1980-2017), maturity (2017-2030), and decline (post-2030). For China, the stages are: introduction (1984-2001), growth (2001-2015), maturity (2015-2023), and decline (post-2023). The differences are notable: China’s growth phase started later but is more rapid, and its maturity phase is shorter, reflecting accelerated development in the China robot industry due to policy support and investment. The logistic parameters for China can be estimated from the data. For instance, using non-linear regression, we derive values for \( K \), \( r \), and \( t_0 \).假设 the saturation level \( K \) for China robot patents is around 15,000, based on extrapolation, the growth rate \( r \) is approximately 0.5 per year, and the inflection point \( t_0 \) is near 2010. This yields the equation:
$$ P(t) = \frac{15000}{1 + e^{-0.5(t – 2010)}} $$
which fits the observed cumulative patent counts. Comparing this to the global model, where \( K \) might be 50,000, \( r \) is 0.3, and \( t_0 \) is 2000, we see that China’s robot patenting is catching up quickly. The shorter maturity phase in China suggests that technological saturation may occur sooner, but this could also lead to earlier decline if innovation stalls. Therefore, continuous R&D is crucial for sustaining the China robot industry’s momentum.
To predict hotspot technologies, we employ citation analysis using tools like Citespace. Burst detection algorithms identify patents with sudden increases in citations, indicating emerging technologies. Based on data from 2010-2012, we forecast that by 2018, key hotspots will include patents related to advanced control systems,焊接 automation, and collaborative robots. These technologies are expected to commercialize by 2025, shaping the future of the China robot market. The burst strength \( B \) for a patent can be computed as:
$$ B = \frac{C_{\text{peak}} – C_{\text{base}}}{\Delta t} $$
where \( C_{\text{peak}} \) is the citation count at peak, \( C_{\text{base}} \) is the base citation count, and \( \Delta t \) is the time interval. Patents with high \( B \) values are likely to be influential in the China robot领域. For example, patents like CN102079089-A and US5337732-A show strong burst patterns, suggesting they may become foundational for next-generation China robot applications.
Based on our analysis, we propose several policy recommendations to enhance the China robot industry. First, universities and research institutes should leverage their innovation优势 to pioneer breakthroughs in core technologies. Their high patent output per entity indicates strong R&D capabilities, but commercialization remains a challenge. Second, fostering collaboration between academia and industry is essential. Enterprises can benefit from academic research to develop marketable products, while academics gain real-world insights. This synergy can be encouraged through government incentives, such as tax breaks for joint projects or funding for technology transfer offices. Third, policymakers should consider the technology life cycle when designing strategies. The rapid growth and short maturity of China robot patents imply that planning cycles need to be adaptive. For instance, extending科技规划 periods from 5 to 7 years could align better with the time required for technologies to move from feasibility to societal implementation, as seen in initiatives like the EU’s Horizon 2020. Finally, large enterprises should be guided to increase their R&D investments and patent布局 awareness. As major players in the China robot market,他们 can drive significant technological advancements and compete globally. Policies could include subsidies for corporate R&D or recognition programs for innovative firms.
In conclusion, our patent map study provides a comprehensive overview of the industrial robot technology landscape in China. Through detailed analysis of patent data, we have identified trends in application years, institutional distributions, technological focuses, and life cycles. The China robot industry is characterized by rapid growth, concentration in key technologies like manipulation and welding, and a dominant role for universities in patenting. However, opportunities exist for diversification and deeper industry-academia collaboration. The logistic model forecasts a mature phase by 2023, urging timely interventions to sustain innovation. Hotspot technologies, such as those in control and automation, are poised to shape the future. By heeding these insights, stakeholders can navigate the evolving China robot ecosystem, ensuring that it remains competitive in the global “Industry 4.0” landscape. Our work underscores the value of patent maps as tools for technological intelligence, offering a roadmap for strategic decision-making in the dynamic field of China robot development.