In the 2025 Spring Festival Gala, a group of robots wearing colorful cotton-padded jackets and performing a “cyber yangge dance” entered the public’s view. It was on this night that the real era of humanoid robot development began. More and more people started to discuss when robots would truly enter our lives.
Currently, the robot industry is transitioning from prototype demonstration to large-scale mass production. Due to the lack of a “mass-production grade” computing platform that can meet the requirements of high reliability, high integration, and high performance, robot manufacturers have yet to break through the bottleneck of large-scale mass production. These are exactly the areas where Black Sesame Intelligence, a leader in intelligent vehicle computing chips, excels.
Black Sesame Intelligence believes that the current state of the robot industry is roughly equivalent to that of the intelligent vehicle industry in 2018, but the former’s growth rate will far exceed the latter’s. Driven by the intelligent vehicle industry, the current chip industry has become very mature. Meanwhile, excellent AI large models such as large language models, multi-modal large models, and world models have emerged. Therefore, Black Sesame Intelligence predicts that the robot industry is on the verge of an explosion.
Based on the above judgment, Black Sesame Intelligence entered the robot track, filling this crucial gap.
Meanwhile, embodied intelligence has reached the crossroads of commercialization. With its chip R & D foundation and supply chain system capabilities accumulated in the automotive industry, Black Sesame Intelligence is undoubtedly a powerful promoter of the commercialization of embodied intelligence. It can be said that Black Sesame Intelligence’s entry is just in time.
Robots, Trapped on the Eve of Mass Production
In 1959, American inventors George Devol and Joseph Engelberger successfully developed the world’s first industrial robot, “Unimate,” which was officially put into use in a General Motors factory in 1961. Since then, the robot industry has completed its initial exploration and officially moved from the laboratory to the industrial application field.
After decades of long – term development, the robot industry is now entering a new stage of “driven by diverse demands and full – scene penetration.”
First of all, with the continuous progress of robot technology, its downstream application scenarios have become more and more diverse. In addition to traditional industrial fields such as the automotive manufacturing and consumer electronics industries, where robots are used for tasks like handling, spraying, and welding, robots are also widely used in emerging fields such as new energy vehicles, photovoltaics, and biomedicine. For example, the charging robot for new energy vehicles can automatically complete the entire process of plugging and unplugging the charging gun.
As the application fields become more extensive, different functional requirements naturally emerge. For example, industrial humanoid robots need to have the abilities of autonomous recognition, planning, and decision – making to complete complex operation tasks. In the long run, household service robots focus more on human – machine interaction, environmental perception, and personalized service capabilities, such as doing laundry, cooking, and chatting with users according to instructions.
Correspondingly, the user groups served by robots have also changed. In the past, the main users of robots were large institutions such as industrial enterprises. Now, with the popularization of robot technology and the reduction of costs, more and more small and medium – sized enterprises, scientific research institutions, and even individual consumers have become users of robots. Different user groups have different demands and expectations for robots, which further promotes the diversified development of the robot industry.
In summary, the mainstream demand in the robot market is shifting from “usable” to “easy to use”: robots are required to have high – level capabilities such as autonomous perception, complex decision – making, safety control, and system coordination. This places higher requirements on the chips used in robots. For example, they need to meet higher standards in terms of computing power, integration, and real – time performance.
Among them, the computing power chip is the multi – dimensional embodied intelligence engine of the robot. Its architecture determines the dimension of spatial intelligence and is the physical basis for the robot to achieve understanding, planning, reasoning, and interaction. The perception, planning, decision – making, and control dimensions required for a robot to complete tasks are much higher than those of intelligent driving, requiring a more multi – dimensional computing architecture for support. The use of heterogeneous computing power in intelligent driving computing platforms has accumulated experience for multi – dimensional intelligent computing, and the computing architecture with multi – form collaborative operation will play a greater role in robot applications.
Currently, the chip capabilities of the robot industry still cannot meet higher expectations. This is also one of the main reasons why the robot industry is trapped in prototype demonstration and unable to move towards large – scale mass production. The experience that Black Sesame Intelligence has accumulated in automotive chips can be reused in the robot field.
The above – mentioned changes are also promoting the evolution of robot systems from lightweight solutions mainly based on MCUs and single – core CPUs to intelligent module platforms with AI reasoning, vision fusion, and heterogeneous computing capabilities.
On November 20th, Black Sesame Intelligence officially launched the multi – dimensional embodied intelligence computing platform – Black Sesame Intelligence SesameX™. This platform includes three major computing platforms: the SesameX Kalos computing platform dedicated to commercial service robots, the SesameX Aura general computing platform for multi – task execution robots, and the all – around computing platform SesameX Liora for the “brain” of embodied intelligence.
The SesameX full – stack computing platform for robots is a complete set of “systematic computing platform from edge – side modules to whole – brain intelligence.” From hardware, software, toolchains to model ecosystems, it is all self – developed by Black Sesame Intelligence. In developing this platform, Black Sesame Intelligence reused its experience and capabilities in designing chips for the automotive market, with three major advantages: reuse of automotive – grade IP, functional safety certification, and an open technology ecosystem.
An obvious trend is that with technological breakthroughs in multiple fields such as AI cognition, hardware execution, and environmental interaction, combined with factors such as the explosion of market demand and the improvement of the industrial ecosystem, the era of whole – brain intelligence has arrived in the robot track. The SesameX full – stack computing platform for robots is the industry’s first computing platform for the commercial deployment of robots. It integrates perception, control, safety, and intelligence and supports the realization of whole – brain intelligence, enabling robots to truly transform from “mechanical executors” to “embodied intelligent partners.”
Shan Jizhang, the founder and CEO of Black Sesame Intelligence, compared the current robot industry to the intelligent vehicle industry in 2018 and predicted that its growth rate would far exceed the latter’s: “On the one hand, the underlying chip technology has become highly mature; on the other hand, the rapid evolution of large language models, multi – modal large models, and world models has laid a solid foundation for the explosion of robots.”
The predictions of relevant institutions also confirm Black Sesame Intelligence’s judgment. The global market continues to grow, especially in the fields of service robots and special robots. It is estimated that the global service robot market will exceed $40 billion in 2025, with a compound annual growth rate of over 20%. Naturally, there is a huge chip market within it.
According to a GIR survey, in terms of revenue, the global industrial robot chip revenue was approximately $2.322 billion in 2024 and is expected to reach $4.374 billion in 2031. According to QYResearch statistics and forecasts, the global market sales of dedicated chips for humanoid robots reached $686 million in 2024 and are expected to reach $3.271 billion in 2031. Calculated in RMB, this will be a market worth 20 – 30 billion yuan. This is also one of the reasons for Black Sesame Intelligence to enter the market.
Entering the Market with “Capital Advantages”
When an automotive chip designer decides to participate in the R & D of robot chips, those in the know understand that this is equivalent to “entering the market with a silver spoon in one’s mouth.”
After all, in the wave of new energy vehicles, automobiles have completed the initial transformation from traditional machinery to intelligence. High – level urban + highway intelligent assisted driving can liberate the driver’s hands to a greater extent and take passengers from point A to point B. Meanwhile, the intelligent cockpit also plays a huge role in enhancing the comfort and convenience of driving and riding.
Black Sesame Intelligence’s advantage in entering the robot track lies in the high overlap between the robot industry chain and the automotive industry chain. The resources such as the supply chain, quality control, and customer ecosystem that Black Sesame Intelligence has accumulated in the automotive field can be seamlessly transferred, forming a strong competitive barrier.
Let’s first look at the technical level. As a complex system integrating multiple sensors, actuators, and AI algorithms, robots are undergoing three key evolutionary trends. The first is multi – modal perception fusion. As robots enter unstructured environments (such as outdoors, business districts, industrial parks, and households), single – vision or radar sensors are difficult to meet the task requirements. The technological trend is moving towards the fusion of multiple modalities such as RGB, depth, IMU, and voice, which places higher requirements on data channel bandwidth, image pre – processing, and AI perception computing power. This trend promotes the transition of robot platforms from the traditional ARM + vision chip structure to a heterogeneous computing chip platform integrating ISP, CV acceleration, and NPU.
Second, the enhancement of edge intelligence and local reasoning capabilities. To reduce cloud dependence, improve response speed and security, robot systems are gradually moving AI reasoning to the local level. In typical scenarios such as commercial services, security, and patrol, they need to have medium – to – high – intensity reasoning capabilities. Therefore, chip modules with (16 – 100+ TOPS) have become a necessity. Especially when dealing with tasks such as target detection, semantic segmentation, and path planning, the NPU architecture with efficient INT8 reasoning capabilities and low – power consumption performance has become the mainstream choice.
Third, the trend of system – level heterogeneous integration is obvious. High – level robot tasks often involve image processing, neural networks, and real – time control requirements. The technological trend requires the chip platform to have high – performance CPU/GPU, dedicated NPU/DSP, and automotive – grade safety control units, and support high – bandwidth interconnections such as PCIe, MIPI CSI, and TSN. This SoC architecture is becoming the core carrier of the robot’s “intelligent brain.”
The Black Sesame Intelligence robot computing platform can exactly meet the module solutions for the above three trends of “multi – modality, heterogeneity, and marginalization.” In terms of performance, power consumption, interfaces, and ecological compatibility, it is gradually approaching international standards and is suitable for various service and special robot deployment scenarios.
An obvious trend is that the robot market is evolving from “manufacturing tools” to “intelligent partners,” and the foundation supporting this leap is the intelligent computing platform integrating hardware and software.
In addition to the reasons mentioned above, the difficulty in cost control and supply chain management are also the constraints that prevent robots from moving towards mass production. Black Sesame Intelligence has significant advantages in supply chain management, cost control, and large – scale production, and can transfer its capabilities from the automotive field to the robot field.
The Second Growth Curve
Black Sesame Intelligence’s entry has opened up another possibility for the robot industry. Robots are the carriers of “embodied intelligence,” and the core is to have a powerful “intelligent central” computing platform that can handle the entire chain of tasks including perception, cognition, decision – making, and control. Therefore, high – performance chips are of crucial significance for the future development of the robot industry.
With Black Sesame Intelligence’s entry, high – performance chips are being applied to robot products, promoting the acceleration of robot mass production and the application of robots in more diversified niche fields. This determines whether robots can grow from “automated machines” to “intelligent agents” with strong self – learning abilities.
The SesameX™ multi – dimensional embodied intelligence computing platform launched by Black Sesame Intelligence this time aims to break the boundaries of robot intelligent perception and motion control, reshape the robot’s intelligent central, and take a crucial step in the evolution of the robot industry from “specialized equipment” to “general intelligent agents.”
For Black Sesame Intelligence, it has found a new growth engine.
Founded in 2016, Black Sesame Intelligence has been focusing on the R & D of automotive – grade intelligent vehicle computing chips and solutions since its inception. It was the first in the industry to launch the automotive – grade high – performance autonomous driving chip A1000 family, which is also one of the most mature, most mass – produced, and most cost – effective single – chip solutions for integrated parking and driving in China at present. It also launched the Huashan A2000 family of computing chips for the next – generation AI models with higher performance and efficiency, and the Wudang C1200 family, the first high – performance cross – domain fusion computing platform in China with the ability of cabin – driving integration and supporting integrated cabin – driving, achieving a technological breakthrough of integrating four chips into one. Currently, Black Sesame Intelligence has entered the leading echelon in the field of intelligent driving chips in China.
However, the rapidly evolving automotive requirements are forcing automotive chip manufacturers to continuously invest huge amounts of money in R & D. If Black Sesame Intelligence can play the robot card well, in addition to obtaining more revenue, the two business lines can also jointly enhance the brand’s influence, and Black Sesame Intelligence will enter the fast lane of dual – wheel drive.
With three major advantages: the accumulation of basic technologies for assisted driving, the automotive – grade safety guarantee system, and the construction of an open technology ecosystem, Black Sesame Intelligence has built a full – chain technical ability of “perception – cognition – decision – making – control,” forming a unique competitive barrier for entering the robot track.
The SesameX platform is the only robot computing platform in the industry that meets automotive – grade safety standards. It can quickly adapt to high – safety robot scenarios. Its release marks a substantial step in Black Sesame Intelligence’s layout in the robot field and will also be a crucial step for Black Sesame Intelligence to find new growth points.
In fact, the market has already confirmed this with real money.
On November 20th, at the product launch event of Black Sesame Intelligence’s robot platform, the two parties signed a strategic cooperation agreement on – site. Joyson Electronics is a global leading provider of intelligent vehicle technology solutions, positioning itself as an “automotive + robot Tier1.” In the robot field, it is driven by two wheels: robot core components and industrial scenarios, supporting the product R & D of global robot partners and has cooperated with many well – known automotive enterprises and robot companies for supply.
Not only Joyson, but as soon as the SesameX platform was launched, it reached cooperation with leading robot enterprises in the field of embodied intelligence, marking the first step in commercialization.
Since then, Black Sesame Intelligence has officially entered a new stage of “automotive – robot” technology collaborative empowerment. This will open the door to a trillion – level new market for Black Sesame Intelligence.
link