Advanced humanoid robot with built in computer for real time application

Abstract

To provide a coordinated control system designed with a small form factor that can be located within a robot or an automated inline manufacturing system, and a coordinated control method.SOLUTION: An AI-enabled processor includes: on board advanced distributed control hardware incorporated within a System on Module (SOM) board, each SOM comprising four processors or nodes; wireless communication to send and receive commands and any other data; a plurality of sensors for perceiving the environment; an external host flashing computer dedicated for uploading / downloading firmware, cloning and configuration setup; a non-volatile memory for storing control algorithms, configuration files and other essential data; a user interface for robot operation, training and setup; an internal star network; and a switch board.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a system and method for increasing the operating speed of robots used in real-time applications by implementing a small form factor Advanced Distributed Computing Hardware System (ADCH) capable of performing parallel computing for multitasking processes. The present invention provides a powerful and efficient system resulting from implementing an end-to-end autonomous standalone robot by incorporating an ADCH into the robot with a digital signal processor board capable of running multiple neural networks in parallel and processing a wide range of data from an array of vision systems, interface devices and sensors, aided by artificial intelligence inference, deep learning and reinforcement learning software algorithms. [Background technology]

[0002] A robot is an electromechanical assembly controlled by one or more computer programs and / or electronic circuits. Autonomous robots can perform desired tasks in unstructured environments without continuous human intervention. In contrast, semi-autonomous and non-autonomous robots often require human intervention in the form of training to perform actions such as loading, unloading, sorting, machining, and packaging specific objects. Robots are used in a variety of fields, including manufacturing, space exploration, pharmaceuticals, surgery, and automotive. Dedicated robots are generally designed to perform a single task or set of tasks, such as moving from one point to another, and to perform multiple tasks along the way or at a predetermined destination before being commanded to perform a new task. Humanoid robots are a category of robots that attempt to emulate predetermined human tasks, including, but not limited to, laborious and dangerous jobs, among many other tasks. Humanoid robots are typically constructed with anthropomorphic characteristics that allow them to understand and interpret human commands through their behavior. Humanoids may be designed for functional purposes, such as human interaction with tools and environments, or for more intelligent tasks using artificial intelligence and continuous machine learning. In some cases, humanoid robots may be equipped with heads designed to replicate human sensory functions, such as eyes and ears, so that they can be programmed to see, perceive, and understand their operating environment, hear commands, and act based on programmed actions.

[0003] Conventional humanoid robots typically do not have embedded computing capabilities similar to high-end computers or servers. Instead, humanoid robots often rely on external computing systems to perform complex tasks and calculations while still performing simple tasks. Conventional humanoid robots are generally composed of a combination of hardware components, sensors, and actuators, along with embedded systems that enable basic processing and control functions. These embedded systems are responsible for handling low-level tasks such as motor control, sensor data processing, and actuator control.

[0004] To become more autonomous and intelligent, humanoid robots are being required to use artificial intelligence (AI) to improve their performance. AI-based humanoid robots are robots designed to interact with humans in a human-like manner. They incorporate artificial intelligence (AI) technology to perceive their surroundings, make autonomous decisions, and perform multiple tasks. Effective collision avoidance strategies for overcoming obstacles often combine multiple approaches, using infrared and proximity sensors, mapping, path planning, and dynamic adjustments to ensure safe, collision-free, and efficient robot movement. Such robots are equipped with a variety of infrared, tactile, proximity, and other types of sensors, including cameras, microphones, and touch sensors, to gather information from their environment. AI algorithms process such sensor inputs to understand their operating environment and make informed decisions. However, processing environmental information requires high-speed analytics, deep learning, reinforcement learning, and machine learning software modules to handle various types of computationally intensive data. Conventional systems typically utilize general-purpose microprocessors located on external servers to process information and transmit results to the AI robot, resulting in time delays. Due to their inherent hardware configuration, command processing is performed sequentially with very minimal or no parallelism. This affects the speed of the robot, leading to reduced productivity and efficiency.

[0005] The AI capabilities of humanoid robots can vary widely. Some robots are programmed with predefined actions and responses, while others can use machine learning techniques to learn from experience and improve their performance over time. Deep learning algorithms, a type of machine learning, are being used to enhance the cognitive abilities of humanoid robots, enabling them to recognize objects and faces, listen, understand, speak, and even express emotions.

[0006] The applications of AI-powered humanoid robots are diverse. They can be used in fields such as medicine to assist with patient care and rehabilitation exercises. They are also used in education, where they can act as interactive companions or tutors for children with special needs. In the field of customer service, humanoid robots can provide assistance and information in public spaces such as airports or shopping malls.

[0007] However, for more advanced calculations and decision-making processes, humanoid robots often rely on external computing resources. These resources can include cloud-based servers or remote computers that handle large computational loads. The robot transmits sensor data to the external computing system, analyzes the information, performs complex calculations or analyses, and sends back instructions or commands to the robot. By leveraging external computing power, humanoid robots can benefit from broader computing capabilities, access to large amounts of data, and the ability to utilize advanced algorithms and machine learning models. This approach increases the flexibility and scalability of the tasks and applications that humanoid robots can perform. However, as humanoid robots are required to perform more intelligent tasks using artificial intelligence and deep learning algorithms, the amount of data to be analyzed begins to increase, and external servers are no longer able to meet the timing constraints. Because all humanoid movements have a significant time lag, they are not suitable for high-speed, real-time applications, such as many manufacturing and non-manufacturing industries, such as contact lens manufacturing, electronic component inspection, sorting, and packaging.

[0008] AI-enabled humanoid robots have made great progress but still face challenges. Robots that can navigate complex environments, interact naturally with humans, and handle unpredictable situations are still in the early stages of development. Nevertheless, ongoing research and technological advances continue to drive the development of more sophisticated and capable AI-enabled humanoid robots.

[0009] The need for massive data processing demands is an ongoing challenge, and therefore, one immediate solution is to provide a collaborative robot control system with high computing power within the humanoid robot to ensure rapid and real-time responses. It is noteworthy that technological advances continue to evolve, and future generations of humanoid robots will require on-board computing power. The miniaturization of semiconductors and the significant improvement in processing power efficiency make it possible to incorporate multiple advanced digital signal processors (DSPs) that can handle dynamically assigned tasks, thereby enabling the design of fast and highly intelligent anthropomorphic robots. Summary of the Invention

[0010] In view of the background art, an objective of one aspect of the present invention is to provide a powerful collaborative robot control technology that includes an on-board high-performance AI computer module, such as an NVIDIA AGX series module, capable of processing large amounts of data from multiple input sources, and executes different tasks in parallel using a dedicated set of processors within the AI computer module, instead of communicating with an external computer, either on the cloud or a server. In particular, this technology can enable a humanoid robot to utilize the on-board processor to implement artificial intelligence and deep learning to execute a set of tasks without relying on external resources that make it difficult to achieve real-time response.

[0011] Another objective of the present invention is to provide industrial humanoid robots with on-board intelligence that can process information, make autonomous decisions, infer motion tasks based on commands, external interface inputs, interpret on-the-fly images captured by a camera, and quickly perform tasks without the need to send information to an offline server and wait for a response. The robot's intelligence is enhanced through the implementation of reinforcement learning, where it receives feedback in the form of rewards or penalties based on the response to each command. Over time, the humanoid robot improves its decision-making process to achieve better results in the form of accuracy, speed, consistency, and reliability, among many other motion capabilities.

[0012] Another object of the present invention is to provide an ADCH system for integration into in-line automation systems associated with specific processes in manufacturing environments where full-scale robotics is not required.

[0013] Another objective of the present invention is to enable humanoid robots to be more autonomous and adaptive, operating in a variety of environments and efficiently performing complex tasks through deep learning techniques. Advanced robots utilize multiple types of sensors, such as cameras, remote sensing units, radar, and transducers, to collect data about their environment. Sensor fusion techniques combine data from such sensors and other external interfaces, allowing the robot to gain a more comprehensive understanding of its surrounding environment, thereby improving its AI responses over time.

[0014] Another objective of the present invention is to enable humanoid robots to be more autonomous and quickly calculate kinematic controls for the robot so that the robot's end effector can move to a specified destination. This is implemented via an on-board computing processor that implements inverse kinematics, ensuring smooth, precise, and jerky-free movements of the robot arm and enabling the implementation of anthropomorphic functions.

[0015] Another objective of the present invention is to embed humanoid robots with multiple dedicated on-board computing processors (e.g., NVIDIA® Jetson AGX Xavier series) so that all data analysis, data manipulation, and other real-time algorithms can be distributed across a parallel system architecture embedded in the robot, ensuring ultra-fast responses to specific commands from module-specific and device-specific accessories.

[0016] The goal is for these robots to be able to respond to their environment and adapt to changing conditions and interactions to perform tasks in real time.

[0017] Another object of the present invention is to enable high speed transfer of large amounts of data (e.g. high resolution images) between different nodes or processors over USB C and Ethernet star networks without the speed limitations of the common BUS that is also used for board-to-board communication.

[0018] Such a high-speed data transfer strategy allows software applications to create maps of the manufacturing environment using sensor inputs. These maps can be used to identify obstacles and incorporate techniques like Simultaneous Localization and Mapping (SLAM) to determine the robot's position and orientation within the mapped environment, thereby planning the shortest collision-free trajectory path. Many of the analyzed trajectory paths may be further stored and utilized as part of predefined actions or movements to perform specific tasks, enabling the robot to navigate autonomously within known indoor environments. The use of multiple processors interconnected via multiple data transfer networks, such as USB-C, Ethernet, and common bus architectures, enables extremely fast data analysis and continuous updating of planned trajectory paths based on real-time sensor data, allowing the robot to adapt to changing environmental conditions and avoid new obstacles as they dynamically emerge.

[0019] Another object of the present invention is to facilitate efficient data transfer for distributed real-time imaging and inference across multiple nodes, eliminating common bottlenecks encountered in conventional multi-GPU server systems that rely on common bus architectures, resulting in slow response and inefficient performance.

[0020] Advances in robotics and artificial intelligence (AI) have made it possible to create humanoid robots with increasingly sophisticated capabilities to perform complex tasks requiring dexterity and precision. These robots often incorporate technologies such as computer vision, natural language processing, machine learning, and sensor systems to perceive and interact with the world around them, making them intelligent anthropomorphic robots.

[0021] As the demand for intelligent robots capable of performing complex tasks increases, intensive on-board computational requirements are necessary to implement such robots in applications such as self-driving cars that can navigate roads and avoid obstacles without human intervention, industrial robots deployed in cleanroom environments, medical robots assisting in surgery, rehabilitation, and other hospital operations that may require extremely hygienic environments to prevent the risk of spreading infection, etc. They may also be deployed in search and rescue operations to locate areas of disaster such as floods, earthquakes, landslides, and other disasters, and agricultural robots for planting, harvesting, monitoring, and packing. [Brief explanation of the drawings]

[0022] The invention will be better understood from the following description of non-limiting embodiments, with reference to the accompanying drawings, in which: [Figure 1] FIG. 1 is a block diagram illustrating an example of a highly distributed computing hardware system that can be used for collaborative robotic control implemented in a humanoid robot. [Figure 2]The diagram shows the configuration of a single-board highly distributed computing hardware system, with each processor equipped with a GPU (graphics processing unit), an ARM-type CPU (central processing unit), a VPA (vision processing accelerator), and a DLA (deep learning accelerator), which handle different tasks and multiple types of interfaces to control and monitor the functions of a humanoid robot. [Figure 3] This is a block diagram of a three-board highly distributed computing hardware system, showing multiple GPUs controlling multiple interfaces via a communication module implemented through the main board and a common BUS network. [Figure 4] 1 shows a floor plan of a single-board highly distributed computing hardware configuration. [Figure 5] A side view of a three-board highly distributed computing hardware configuration is shown. [Figure 6] Figure 5 shows an isometric view of a three-board highly distributed computing hardware system. [Figure 7] FIG. 1 shows a front view of a humanoid robot incorporating advanced distributed computing hardware. [Figure 8] An isometric view of a humanoid robot incorporating advanced distributed computing hardware.

[0023] The drawings are not necessarily to scale and may be illustrated by schematic and fragmented representations. In certain cases, details that are not necessary for understanding the embodiments or that obscure other details may be omitted. DETAILED DESCRIPTION OF THE INVENTION

[0024] We describe an Advanced Distributed Computing Hardware (ADCH) system implemented through peer-to-peer communication between a cluster of processors constituting an action server. Each processor includes a graphics processing unit (GPU), an ARM-based central processing unit (CPU), a vision processing accelerator (VPA), and a deep learning accelerator (DLA) for performing various tasks. In particular, the ADCH system may enable the cooperative processing of tasks assigned by the master processor 100 (FIG. 3) to a network of processors (101-111 in FIG. 3) residing in at least one server through broadcast messaging and service calls to achieve real-time functionality for the robot. Through parallel processing and preemptible services, the network of processors can accomplish complex, multifaceted tasks, such as various cooperative human-like tasks, or even simple cooperative tasks in real time. These preemptible tasks and many other established processes are stored in non-volatile memory within the robot and further refined and enhanced using machine learning and deep learning algorithms.

[0025] One embodiment of the present invention uses a set of boards containing a cluster of 12 processors (e.g., NVIDIA's Jetson AGX series) 100-111, as shown in Figure 1. One Ethernet port from each processor is routed to an Ethernet switch 26 to form a star network topology. A peer-to-peer network is established between processes running on each pair of processors (100-111 in Figure 1) in the cluster, communicating using different modes such as broadcast messages, services, and action servers.

[0026] Broadcast messages can be issued by any process in the network without knowledge of the message's subscribers - this is the typical many-to-many connection used for continuous data flow.

[0027] Services are implemented by short-lived remote service calls that are executed sequentially, also known as synchronous service calls. While a remote service call is executing, the scheduled call remains active and cannot be preempted by another remote service call. Thus, a typical service thread remains dedicated to a single service call until it completes execution.

[0028] The action server and client establish a tight coupling of two or more processes to perform different tasks depending on the assigned server, with the option to provide feedback during and upon completion of a service call or request from the client. Note that the action server is designed to be preemptible and non-blocking, which means that it can perform multiple tasks.

[0029] All processes are designed to be fine-grained and modular, with each process performing a well-defined task and having callable interfaces that are exposed by the process to one of the aforementioned communication modes (broadcast messages and remote service calls) based on requirements.

[0030] High-reliability safety systems are integrated into the ADCH. These include collision detection using infrared and proximity sensors, emergency stop mechanisms, and fail-safes to prevent harm to the robot or its surroundings, as well as self-diagnostics that detect and report abnormalities in real time within the computer and inappropriate responses or failures from external hardware interfaces. Furthermore, password-protected data security and privacy are ensured within the ADCH, allowing data to be processed and stored locally without the risk of data being leaked during transmission to external servers, especially when sensitive applications are involved.

[0031] The modular processes are distributed across all 12 processors (100-111 in Figure 1), with one master processor 100 able to dictate which processes are assigned to which processors. The master 100 is connected to the display and mouse 10 in Figure 1 via an HDMI port to train, configure, program, and invoke the tasks required to perform actions. Given the open nature of the processes and network topology, the master has the ability to dynamically assign specific processors to specific processes depending on the required processing power and processor load at any given stage. In effect, the master manages the cluster much like an operating system (OS) scheduler manages a multi-core processor.

[0032] Referring to FIG. 1, the ADCH system 200 includes three boards B1, B2, and B3, each consisting of four Xavier processors. A cluster of a total of 12 processors (100 through 111 in FIG. 1) is available for executing multiple tasks in parallel through high-speed communication over a dedicated PCI bus on the main board 30. Due to space constraints, not all Xaviers are shown in FIG. 1. In the embodiment shown in FIG. 1, reference numeral 100 is a master that manages process allocation among processors 101 through 111. External interfaces, such as motor 20 via motor driver 18, cameras 16 and 14, and sensors (not shown), are connected to another processor 103 in FIG. 1. All Ethernet and I2C connections from each processor are terminated on the BUS for connection to the main board 30 via BUS connector 22. Inter-board communication may be enabled to operate peer-to-peer or over a high-speed bus interconnecting the three processor boards B1, B2, and B3. The main board 30 also provides general input / output connections 23 via an I2C I / O expander 24 and communication to the Internet 28 via an Ethernet switch module 26. Power is provided by 21. Due to the high speed data required by the cameras 14 and 16, they are interfaced to USB C ports that provide peak speeds of up to 40 Gbps, enabling real-time data transfer, especially when using high-resolution cameras. A USB A port is utilized when data transfer rates below 10 Gbps are sufficient for the connected accessories. 3BUT are I / O ports to which sensors and / or switches may be connected (not shown) to provide operational status feedback, hard reset, and other hardware I / O configuration options that may be required.

[0033] The ADCH system 200 operating system is programmed to configure the 12 processors to perform different types of tasks. For example, in the embodiment shown in FIG. 1, processors 100 and 101 are assigned as control nodes, processors 102 and 103 are assigned as planning and sequencing nodes, respectively, processors 104 and 105 are assigned as user interface and debug nodes, and the remaining processors (processors 106-111) are assigned as imaging nodes. The ADCH system software application provides the flexibility to dynamically change node assignments depending on processor load, accelerating data processing and achieving advanced real-time performance. The system is also designed to scale up performance by adding more processors or reallocating processors to various tasks.

[0034] The ADCH system is embedded in the robot, and all operations are distributed through nodes operating in a multitasking environment that provides extensive parallel processing. The nodes, also called processors, are each connected to nonvolatile memory, allowing for the storage of specific sequences of state conditions. A properly coordinated control mechanism using a star network allows for easy restoration of the robot's operation in the event of a shutdown or power outage. Traditional control systems require a lengthy recovery process that begins with communication with an external server, waiting for the machine's or robot's last known state, restarting the robot from a specific home position, and then executing a command to restore the machine or robot to its last known position. The nodes may store all or any configured data analysis results (visual inspection), which can be used by algorithms to build an artificial intelligence database and enhance machine learning. Nodes 100 and 101 function as control nodes. These nodes assign tasks, track the state of all other processors and their respective computing loads at any given time, and deploy neural networks through any of the other free nodes 102 to 111. Due to the real-time nature of robots, a feature is implemented that allows software applications to control processes performed by either software or hardware, ensuring fast response. The nodes also manage the system's power consumption by turning off the clocks of unused nodes, effectively reducing the robot's heat dissipation. Nodes 102 and 103 manage the planning and sequencing of the ADCH external interface, respectively, with full-duplex capabilities, ensuring optimal use of computing power. The planning and sequencing control nodes help the robot move along the smoothest and fastest trajectory path through optimal calculation of the robot's joint angles and optimized trajectory planning to maintain balance and avoid collisions with intermediate obstacles. Nodes 104 and 105 control the user interface (UI) and debug operations, respectively, during robot training and configuration.Nodes 106 to 111 are dedicated to visual analysis, including imaging, trajectory planning, balance control with optimal robot joint angles, inverse kinematics algorithms, and processing and feedback of results to the network of nodes and the master node 100. The ADCH's master node 100 can optionally assign any node for audio assistance, where the robot recognizes audio commands (voice recognition), speech recognition, natural language processing, and decision-making, enabling the ability to understand natural language commands and generate appropriate natural language responses. Machine learning and deep learning techniques are often employed to improve the robot's capabilities in these areas and perform a task or set of tasks accordingly. Furthermore, feedback in the form of audio responses allows the robot to be scalable and flexible to adapt to various applications. Advanced audio capabilities are implemented when the robot needs to understand and analyze multilingual commands, and they can be implemented in various countries without the user needing to use a specific language UI (user interface) to operate the robot.

[0035] FIG. 2 shows a typical board layout of a single computing board B1, also called a "system on module (SOM)," containing four Xavier processors or nodes. External interfaces terminate on one side of the board and communicate with a main board 30 via connector 22. Some of the interfaces relevant to the present invention are onboard Wi-Fi, Bluetooth, general-purpose I / O, Ethernet, or other communication interfaces for interacting with humans or other devices. This connectivity allows them to receive commands and transmit information. General-purpose IO (GPIO) is provided via an Ethernet communication interface and an I2C interface board 24 via an Ethernet switch board 26. Interface 12 is a host flashing system that plays a key role in configuring the ADCH system by enabling functions such as firmware upload / download and configuration and upload / download of all other related operating parameters. The host flashing computer 12 plays a key role in configuring the ADCH system integrated with the software development kit modules, adding / modifying the operating system kernel of the processor(s) and customizing software applications, boot loaders, and device drivers to perform specific sets of tasks for processing different products, providing scalability and flexibility to adapt the ADCH to changing robot applications. The hot flashing computer 12 may also be used to flash other ADCH systems that control other robots performing the same set of operations, through a process called mirroring or cloning, so that the optimized program of said robot can be copied and cloned to operate another robot, ensuring a stable and consistent operating environment.

[0036] Figure 3 shows a block diagram of a typical ADCH system, including three computing boards B1, B2, and B3 attached to a main board 30 via slot connectors 22. Figure 3 also illustrates how common signals are interfaced between multiple processors, such as 100-111, to achieve a powerful and effective system for monitoring and controlling various devices, such as Optispec cameras 16 and 19, a printer 13, a BO module 15, a robot head camera 14, and a motor controller called Elmo 18. The main board accesses the Internet via an Ethernet switch module 26 and I / O signals via an I2C GPIO expander 24 to communicate with external devices. The configuration shown in Figure 3 is designed with a small form factor that allows it to fit into a robot with sufficient data processing power and minimal need to communicate with an external server for its computing needs. A scalable robotic environment is possible with such a distributed, self-contained system.

[0037] Figures 4, 5 and 6 show top, side and isometric views of an ADCH system that can be easily placed within a robot.

[0038] 7 and 8 show a front view and an isometric view of a robot incorporating an ADCH in this embodiment of the invention. Various devices mounted within the robot are not shown as they are outside the scope of the invention.

[0039] With reference to the illustrated drawings, specific language has been used herein to describe the present invention. However, it will be understood that no limitation of the scope of the technology is intended thereby. Alternatives and further modifications of the features of the control system shown herein should be considered within the scope of this description.

[0040] However, it will be recognized that the techniques can be practiced without one or more of the specific details, or with other methods, hardware components, interface devices, etc. To avoid obscuring aspects of the invention, well-known modules, such as loading and unloading processes or operations, have not been shown or described in detail.

[0041] The subject matter defined in the appended claims is not necessarily limited to the specific features and acts described above. Rather, the specific features described above are disclosed as example forms of implementing the claims. Many modifications or arrangements may be devised without departing from the spirit and scope of the described invention.

Claims

1. 1. A collaborative control system designed in a small form factor that can be placed within a robotic or automated in-line manufacturing system, comprising: On-board highly distributed control hardware configured such that the artificial intelligence enabled processors are embedded within "System on Module" (SOM) boards, each SOM comprising four processors or nodes; Interfaced cameras, motors, general purpose I / O via I2C expander, audio interface, internet, wireless communications such as Wi-Fi and Bluetooth to send and receive commands and any other data; a plurality of electrically connected sensors for perceiving an environment and collecting data from infrared, tactile, proximity, and other types of sensors; An external host flashing computer dedicated to firmware upload / download, cloning, and configuration setup; non-volatile memory for storing control algorithms, configuration files, and other important data; an HDMI-based user interface for robot operation, training and setup; An internal star network of USB-C or Ethernet type for high-speed communication that bypasses the standard PCI bus interface BUS; an Ethernet switch board that allows multiple boards to access Ethernet for both internal communication within the star network and external Internet access; A cooperative control system comprising:

2. an artificial intelligence-enabled system for locally processing data, the artificial intelligence-enabled system enabling significant reduction in latency for making autonomous decisions assisted by artificial intelligence-based machine learning and deep learning algorithms; a user interface for interacting with the robot, the HDMI type user interface comprising a mouse and a display for visual and audio communication with a human operator; a plurality of processors, each of which comprises a GPU (graphics processing unit), an ARM-type CPU (central processing unit), a VPA (Vision Processing Accelerator), and a DLA (Deep Learning Accelerator) for analyzing the environment and achieving real-time performance; The cooperative control system of claim 1 further comprising:

3. Power management functionality within the ADCH to optimize power or battery usage by efficiently managing the processor and disabling unused processor clocks to ensure low heat dissipation The cooperative control system of claim 1 further comprising:

4. A self-diagnostic system that detects and reports in real time any abnormalities within the computer and any faults or improper responses from external hardware interfaces The cooperative control system of claim 1 further comprising:

5. Data security and privacy features that allow data to be processed and stored locally without the risk of data leakage during transmission, which can be important for sensitive applications The cooperative control system of claim 1 further comprising:

6. Dedicated hot-flashing computer for modifying or enhancing the operating system kernel of multiple processors to customize software applications, boot loaders, and device drivers for improved scalability and flexibility The cooperative control system of claim 1 further comprising:

7. The cooperative control system of claim 1 further enabling a distributed, self-contained robotic environment that is scalable and adaptable to new applications.

8. 1. A collaborative control method for a robotic or automated in-line system in a manufacturing environment, comprising: using an embedded computer within the ADCH to coordinate and perform tasks related to the production process, including material handling and quality control; receiving external production commands via Wi-Fi, Bluetooth, or Ethernet from a central manufacturing control system or a software application resident on said embedded computer; Including, the rapid communication within the ADCH is achieved by broadcast messages that can be issued by any process within the star network without knowledge of the subscribers of the messages, resulting in a typical many-to-many connection for continuous data flow; a peer-to-peer network is established between processes executing on each pair of processors; The management of synchronous service calls is implemented by short-lived remote service calls that execute continuously, remain dedicated and active while executing, and are not preemptible by another remote service call; Increased scalability and flexibility with the option to extend the ADCH's capabilities and speed by reallocating unused nodes or processors; Dynamic allocation of nodes or processors by the master node to efficiently distribute tasks for maximum computing speed during data and image analysis during normal operation, debugging, and use of the user interface during training and setting configuration; Establishing a tight coupling of two or more processes for performing different tasks by the action server and the client with the assigned server, with the option to provide feedback during and at the end of the remote service call or service request from the client; A cooperative control method in which the action server is designed to be preemptible and non-blocking, allowing it to perform multiple tasks within the ADCH with password-protected data security and privacy features to process and store data locally without the risk of data leakage during transmission to an external server when used in sensitive applications.

9. 9. The method of claim 8, wherein the processes are designed to be fine-grained and modular to ensure that each process performs a well-defined task with a callable interface.

10. The method of claim 9 , wherein the callable interface is exposed by the process based on requirements for all communication modes (broadcast messages and remote service calls).

11. The method of claim 8, wherein a sequence of robot movements is analyzed and calculated to manage joint angles of the robot to maintain balance.

12. 9. The method of claim 8, wherein the robot's trajectory path is planned through the implementation of an inverse kinematics algorithm to ensure smooth, jerky-free, and accurate movement, allowing for the implementation of anthropomorphic functions.

13. The cooperative control method of claim 8, wherein predefined behaviors or movements are utilized to perform specific tasks for autonomous navigation of the robot in a known indoor environment.

14. 9. The cooperative control method of claim 8, wherein effective AI algorithms are implemented for object recognition, speech recognition, natural language processing and decision making to understand natural language spoken commands and generate appropriate natural language responses.

15. The collaborative control method of claims 8 and 14, wherein the vision system supports understanding of the manufacturing environment and quality inspection capabilities complemented by artificial intelligence reasoning, deep learning and reinforcement learning for efficient end-to-end autonomous applications.

16. 9. The cooperative control method of claim 8, wherein efficient data transfer is facilitated via Ethernet and USB star networks and interfaces within the ADCH to enable distributed real-time image processing and inference across multiple nodes and overcome common bottlenecks encountered in conventional multi-GPU server systems that rely on common bus architectures.

Images