While the embodied intelligence industry is still competing for Demo fluency, model parameters and simulation effects, Siling Robotics has used a blockbuster acquisition to provide the answer to the real industrial world.
Recently, global intelligent robot unicorn Agile Robots announced the completion of its acquisition of ThyssenKrupp Automation Engineering Company. This German industrial automation company with 75 years of engineering experience will be integrated into the Agile system under the new brand “Krause Automation”.
Its engineering capabilities, global network and customer ecology have become the key pieces of the puzzle for Siling Robot to complete project implementation, complete factory delivery and global service capabilities, helping Siling Robot to expand sales channels, quickly penetrate the global high-end manufacturing market, and deepen cooperation with leading companies.
This also marks the beginning of Siling Robot’s end-to-end intelligent factory-wide service.
Siling Robotics, which was founded only 8 years ago, has successfully completed the integration of established industrial forces with its full-stack self-developed technology, massive real machine data and mature commercialization capabilities. It not only broke the industry dilemma of “strong algorithm, weak implementation” and “hot concept, difficult delivery”, but also effectively promoted embodied intelligence from concept to real industrial implementation.
In November 2025, ThyssenKrupp Group disclosed for the first time that it would sell its automation engineering department to Siling Robotics; in April 2026, this transaction affecting the global industrial automation landscape was officially completed.
For the industry, this is not only a model of cross-border cooperation between enterprises, but also redefines the next generation of industrial automation solutions.
Thyssenkrupp Automation Engineering is an international engineering company specializing in the automotive and new energy fields. Its core business covers precision automated assembly and testing production lines for key components of powertrains such as engines and gearboxes. It has accumulated a “thick” experience in automotive production line planning, system integration, and project delivery. It serves the world’s top OEM manufacturers and is a typical representative of German manufacturing’s “heavy engineering, heavy delivery, and heavy stability”.
Its value does not lie in a single device, but in turning technology into factory-level capabilities that can be implemented, deliverable, and replicable – this is one of the core links that are generally missing in current AI robot companies.
In the past, most robot companies only solved the problem of “making the machine move”, while manufacturing customers need a complete set of closed-loop capabilities from solution design, production line planning to system integration, and long-term operation and maintenance.
Through this acquisition, Siling Robot has directly acquired three core assets: engineering capabilities, global network and customer ecosystem. 75 years of industrial production line implementation experience has completed the key link from AI technology to industrial implementation. The newly added top ten business sites in Europe and the United States and hundreds of industry experts have helped it quickly penetrate the global high-end manufacturing market. At the same time, deepening cooperation with the world’s leading automotive OEMs can also further empower the application of new energy vehicles, consumer electronics, medical care, logistics and other scenarios.
Siling Robotics was founded in 2018, with the German Aerospace Center (DLR) as its technical support. In the past eight years, it has built full-stack self-research capabilities from core components to robot bodies, from software platforms, robot large models to system solutions. It is one of the few companies in the industry that masters hardware, algorithms, data and commercialization capabilities at the same time.
In terms of hardware, Siling Robot has formed a complete matrix, including dexterous hands equipped with torque sensors and fingertip tactile sensors for each joint; Agile ONE full-size industrial humanoid robot that can complete micro-screw assembly, power tool operation and other multi-scenario operations; intelligent humanoid robot H20 and wheeled humanoid robot H10 for AI scientific research, education, business and consumer fields; Diana, Thor, Franka covering precision assembly, flexible processing, medical and health care and other scenarios.
Force-controlled robotic arm series such as FR3; fully automatic composite robots that integrate logistics and operations; and self-designed core components such as motors, reducers, vision/force/tactile sensors, control the underlying hardware, and support the implementation of automation in multiple industries.
In terms of software, Siling Robot independently developed the AI-driven software platform AgileCore.
The platform can realize hardware interconnection, algorithm interoperability, and user collaboration, break the barriers of equipment fragmentation, and support natural language programming, one-click line change, and cross-production line reuse.
Siling Robot has its own Robot Foundation Models. The construction and evolution of this model is a complete training system that integrates real industrial data, simulation environment data and remote operation teaching data.
Based on this, Siling Robot’s bipedal humanoid robots and wheeled humanoid embodied intelligent robots can transfer capabilities between different task scenarios to achieve more robust and safe complex operations.
All hardware products of Siling Robot can be uniformly scheduled by the AgileCore AI software platform. As a technology hub, the platform allows Agile ONE to seamlessly collaborate with the company’s existing dexterous hands, force-controlled robotic arms, fully automatic composite robots and other products to form an intelligent production system covering the entire process of “perception-decision-execution”.
Siling Robots gives full play to its rich intelligent robot hardware platform and application ecological advantages, and its services cover many AI leading research institutions and universities around the world. For example, it has reached cooperation with Google Deepmind on the application of artificial intelligence in robots, and has realized the sales of robot bodies.
Siling Robot continues to combine the capabilities of these base model head mechanisms in large language models, visual models and autonomous motion models to explore the possibility of embodied intelligent robots moving from the laboratory to practical applications.
The truth of the embodied intelligence industry is: being able to run a demo is not a skill. Being able to deliver stably, sustain profits, and replicate on a large scale is the core competitiveness.
Siling Robotics has doubled its revenue for seven consecutive years, delivered more than 20,000 sets of robots and system solutions to global customers, and served many Fortune 500 customers, proving that its technology has completed the closed loop from concept to business.
The commercialization of Siling Robot focuses on high-value industrial scenarios and has created multiple industry benchmark cases. Among them, Xiaomi’s fully unmanned packaging production line for mobile phones uses AgileCore to achieve multi-device collaboration, support rapid switching of multiple product lines, and achieve UPH
600, the utilization rate is 85%, and the efficiency is far higher than that of similar manual production lines in the industry; the automation rate of Xiaomi’s automotive electronic domain controller assembly line has been greatly improved, and the output per hour is 30% higher than the industry average, significantly reducing labor costs; the mobile automatic receiving robot in BMW’s logistics automation solution realizes the intelligentization of the entire process of receiving incoming materials and tallying, effectively solving problems such as difficulty in manual scanning, error-prone, and high labor intensity.
The commonality of these cases is that they do not replace single-point equipment, but provide whole-plant-level flexible automation solutions to truly help customers reduce costs, increase efficiency, and improve quality. This is also the core capability that Siling Robot will continue to strengthen after the acquisition of thyssenkrupp Automation Engineering.
Relying on strong engineering and product capabilities, Siling Robot has broken through the boundaries of traditional industry.
In the field of AI, it helps many of the world’s top institutions to provide data collection services (data farms); in the medical field, Diana7 Med medical robots are used in orthopedics, neurosurgery, rehabilitation physiotherapy, etc., and cooperate with top tertiary hospitals; in the field of energy and power, it launches an unattended solution for smart high-voltage power distribution rooms to solve the problems of complex and hazardous environment operations; at the same time, it continues to expand in the fields of logistics, agriculture, animal husbandry, education, etc., forming a commercial pattern of “industrial-based, multi-scenario collaboration”.
As of now, Siling Robotics has more than 3,200 employees worldwide. It has one of the largest R&D teams in the field of AI and robotics. Its business covers major markets around the world. It is one of the few technology companies that can compete head-on with international industrial robot giants.
The biggest difference between physical AI and digital AI is that digital AI can rely on public data to close the loop, while physical AI must rely on real physical interaction data iteration.
For example, autonomous driving requires large-scale road testing, and the evolution of physical AI cannot be separated from the application of tens of millions of robots in real scenarios.
One of the core barriers of Siling Robots is massive, high-value, and exclusive real machine data.
At the data level, Siling Robot has built a four-dimensional system covering real machine data, simulation data, human-computer interaction data, and Internet data. Real machine data determines the lower limit of intelligence, and Internet-level data supports breakthroughs in the upper limit of intelligence.
At present, it has deployed more than 20,000 robots and system solutions around the world, and has accumulated a massive collection of real industrial physical interaction data, covering the entire process of assembly, inspection, logistics, human-machine collaboration, etc. This data comes from automobiles, 3C, medical and health care and other scenarios, including force control, touch, vision, multi-machine collaboration and other information. It supports the improvement of model generalization capabilities and can be directly used for robot training to better meet actual application needs.
More importantly, Siling Robot has formed a complete data flywheel. It trains the model through real machine data, and further expands the deployment after the model optimizes the product, thereby generating more data, forming a self-reinforcing closed loop, making Siling Robot’s products smarter and more stable the more they are used.
The current industry generally faces the problem of data scarcity. Most companies rely on simulation and synthetic data or a small amount of test data for training, resulting in poor model implementation.
With large-scale commercial deployment, Siling Robot owns the industry’s most scarce real industrial data assets. This is a core advantage that is difficult to catch up in the short term and is also a key support for its physical AI strategy to continue to lead.
Competition in the robot industry has never been a competition of a single algorithm or hardware, but a comprehensive competition of “hardware + AI basic model + scene data + large-scale deployment”.
Siling Robots has spent nearly 8 years completing the leap from technology research and development to product implementation, from single point delivery to whole factory solutions, and from a single market to global layout. By integrating Krause Automation’s 75 years of engineering experience, it has completely opened up the last mile of AI technology and industrial manufacturing.
At the critical juncture of embodied intelligence moving from the “video era” to the “implementation era”, Siling Robot has used pragmatic actions to prove to the industry: true embodied intelligence starts with algorithms, becomes data, is based on engineering, and finally comes to fruition.
In the future, as Krause Automation expands into consumer electronics, medical care, logistics and other fields, Siling Robotics will further consolidate its leading position in global automation solutions, continue to promote the physical AI strategy, and truly integrate artificial intelligence into the physical world.
