TW202303390A - Modular frame for an intelligent robot - Google Patents

Abstract

The invention is a modular frame for an intelligent robot which includes a base and one or more devices for performing specified functions. The base controls the actions and functions of the robot and contains a memory of operating instructions for a plurality of modules, each module performing unique functions. The base has a smart connector. The devices for performing specified functions have a smart connector which contains a unique code for that device or module and firmware for operation of the module. When a module is affixed on the modular frame, the smart connector of the module electronically communicates with the smart connector of the base, thereby providing the base with sufficient operating information to operate the intelligent robot.

Description

A modular framework for intelligent robots

Cross References to Related Applications

This application claims priority to US Provisional Patent Application 63/216,539, filed June 30, 2021, and US Provisional Patent Application 63/233,837, filed August 17, 2021, both of which are incorporated herein by reference.

The present invention relates to robotics, and in particular to modular robots that allow the interchange and/or rearrangement of individual functional modules.

Robots are now ubiquitous and serve a myriad of purposes in many different industries.

The service and personal robotics industry currently faces many obstacles. Including: • It takes an average of 3.5 years and $35 million to develop a mobile autonomous service robot. • The cost of service robots is out of reach for most businesses. • Service robots are designed for specific applications and are not configurable. • The robot is only fully functional when connected to the Internet. • Most robots are reactive, thus limiting their applicability.

Accordingly, there is a need in the industry to make robots affordable and readily available for all businesses and applications; and, there is a need in the industry and domain for devices that efficiently and effectively allow the interchangeability and/or migration of functional modules of robots or system.

To achieve these goals and others, the devices herein facilitate the interchangeability and/or migration of functional modules of robots.

Accordingly, to achieve these and other objectives, the invention disclosed herein is a modular framework for intelligent robots, comprising: a base that controls the motion and function of the robot, and contains operational memory of instructions, each module performs a unique function, and the base has a smart connector; one or more devices or modules, which are used to perform specified functions and have a smart connector, the smart connector contains a unique code for the device or module and firmware for operation of the module; and wherein the smart connector of the module communicates electronically with the smart connector of the base when the module is coupled to said modular frame , so as to provide sufficient operating information for the base to operate the intelligent robot.

Currently, robots are typically developed from scratch, one model at a time, for a specific deployment. For each new deployment, new hardware and software are developed. Such methods have proven to be expensive, time-consuming and inefficient. To address the needs of the cost-sensitive and broadly diverse social robotics market, applicants have created proprietary technology that implements the robotic core needed to develop custom social robots for any need at a fraction of the time and cost of conventionally designed robots Typically 80% redeployment of the "4-pillar" architecture.

The modular framework disclosed in this paper is designed to be flexible, modular and extensible without compromising performance, efficiency and reliability. Such a "Lego" type system ensures that almost unlimited configurations can be built to meet customer deployment goals by integrating the appropriate sensor set for a wide range of applications and designing task-specific external forms.

The modular robot disclosed in this paper is a next-generation M-RAP proprietary modular robot architecture platform, which will further reduce the time and cost of service robot development. All M-Rap core and accessory modules have built-in smart connectors that contain the "DNA" of each hardware element, allowing the robotic "brain" to automatically recognize all hardware elements used or added later, and then automatically Compile the necessary software/firmware to be able to manage and control each robot respectively. This enables virtually unlimited customization for any vertical market deployment. It allows an open plug-and-play API and development kit for third-party accessory manufacturers to develop and sell value-added accessories that can be plugged into each robot in the aftermarket.

The base includes the robot's basic operating system and hardware code that controls and directs the physical devices and modules coupled to the robot's modular framework. In some embodiments, the base includes a CPU, an MCU and a power board, a robot operating system command library, and an artificial intelligence engine. The base or brain contains the communication protocol. Means for autonomous movement may be included.

A modular kit for generating intelligent robots is also provided. The modular kit includes multiple optional base modules and multiple optional functional modules. The base module includes: robot firmware, which includes low-level hardware code that controls and directs at least the base module; a robot operating system (OS), that includes a number of operating instructions for a number of functional modules; a database , which includes the firmware, features and functions of the base module and the function module; and at least one of the smart Lync hub and the smart connector, at least one of the smart Lync hub and the smart connector can be simultaneously connected to multiple a smart connector. A plurality of selectable function modules perform specified functions, each selectable function module having a smart connector containing a unique code for the function module and a code for the operation of the selectable function module firmware.

In addition, according to a preferred embodiment of the present invention, the multiple base modules also include smart firmware, communication protocol, CPU, MCU, power board, robot operating system command library, artificial intelligence engine, memory, and communication modules. at least one.

In addition, according to a preferred embodiment of the present invention, the functional module operates at any designated position and/or designated direction within the intelligent robot.

In addition, according to a preferred embodiment of the present invention, the functional module electronically communicates with the base module via the smart connector.

Furthermore, according to a preferred embodiment of the present invention, the functional module is a drug dispenser, a gripping arm, a torso element, a medical device or device, a sanitary device or device, a transport device or device, a cooking device or device, an entertainment device or device and At least one of personal health and assistive devices.

In addition, according to a preferred embodiment of the present invention, the smart connector includes a unit for providing sufficient operating information to the base module to allow the kit to operate as a smart robot.

In addition, according to a preferred embodiment of the present invention, the smart connector includes at least one of a wireless communication unit and a wired connection.

Furthermore, according to a preferred embodiment of the present invention, the smart connector includes a unit for transmitting the attached module identification, the attached module attachment method, the attached module position, the attached module At least one of a group-specific function, an attached module purchase/prepaid optional function, an attached module-associated design document, and an attached module verification document.

Furthermore, according to a preferred embodiment of the present invention, the associated design material is at least one of a logo, a theme, and a design.

Finally, according to a preferred embodiment of the present invention, the authentication material is licensed, unlicensed and pirated.

In the following detailed description, numerous specific details are set forth in order to provide a thorough understanding of the present invention. However, it will be understood by those skilled in the art that the present invention may be practiced without these specific details. In other instances, well-known methods, procedures, and components have not been described in detail so as not to obscure the present invention.

In a basic embodiment, the invention constitutes a device facilitating the interchangeability and/or migration of functional modules of a robot.

Referring now to Figures 1 and 2, there is illustrated an embodiment of the present invention having a modular kit or framework from which an intelligent robot is generated. Figure 1 shows the basic architecture of the invention, and Figure 2 shows how the modules of the kit are connected.

The kit includes a base module 102 , a plurality of functional modules 104 , a plurality of accessories 106 and a smart lync hub 110 . Each base module 102 and each function module 104 may include a smart connector 108 (shown as X-UIP in FIG. 2 ) for connection to a hub 110 which in turn may provide the base module 102 and the communication between the function module 104.

The base module 102 may control the actions and functions of the resulting robot, and may contain memory for operating instructions for the functional modules 104, each of which performs a unique function.

Each smart connector 108 contains a unique code for its function module 104 or base module 102 and firmware for the operation of the modules. When either module 102 or module 104 can be coupled to smart hub 110, the module's smart connector 108 communicates electronically with smart hub 110, which in turn communicates with base module 102, thereby providing sufficient power for the base. Manipulate information to operate intelligent robots.

Functional modules 104 may include, for example, arms, torso elements (upper, middle, and lower), claws, maintenance equipment, cleaning equipment, medical equipment, cooking equipment, personal aids, fitness equipment, cabinets, drawers, tray sets, etc., some of which Illustrated in Figure 1. Each of these mods and their associated accessories may be manufactured in-house, or exclusively commissioned by an external or third-party individual or other company for use by the individual company (by licensing or other monetary or non-monetary transaction). Accessories 106 may include trays, cup holders or keyboards that connect via proprietary or standard interfaces and do not have smart connectors. In its preferred embodiment, the base modules 102 may be connected individually or may be connected to each other and housed together in a base housing 120, thereby forming a base for controlling the actions and functions of the robot. Base module 102 may include all elements necessary to operate and control the robot. A typical base module 102 and base housing 120 are shown in FIG. 4 .

The base module 102 may include the robot's basic operating system and hardware codes to control and direct the physical equipment. In some embodiments, the base housing 120 can house the base module 102, CPU, MCU and power board, robotic operating system command library, and artificial intelligence engine. Accessories such as wheels, lidar units, keyboards, displays, and USB connectors may also be attached to the base housing 120 . It should be understood that any function, accessory and any functional module can be placed in any position and in any orientation within the generated robot.

The base module 102 may include a firmware library and smart firmware for controlling the smart lync hub 110 . This can include memory, along with the CPU, MCU, power supply, communication modules, and all other electronics needed to operate the robot.

According to the present invention, generated robots can operate offline or online, as described in applicant's co-owned US Provisional Patent Applications 63/304,621 and 63/310,120. For online operation, one or more base modules 102 may include a suitable wireless communication unit. For example, it could be a Wi-Fi or cellular data unit. This may enable it to communicate over the Internet or local area network, or even over cellular telephone service, to receive operating instructions.

More specifically, the intelligent robot can include human-machine interface devices such as speakers, microphones, and touch screens as needed. These may be part of any suitable functional module 104 . In this manner, a person in close proximity to the robot can communicate with the base module 102 to provide instructions regarding specified functions or operations to be performed by the robot.

In the exemplary embodiment, base housing 120 may house one or more smart connectors 108 or smart hubs 110 , also known as smart Lync connectors.

An exemplary embodiment of a skeletal framework of a basic robot is shown in FIG. 3 . In this embodiment, the modules may be joined together as a "skeleton" or "unfinished" module, which may be made of any suitable material, such as metal.

Thus, a user may select one of the plurality of head modules 104A and may attach it to a selected one of the torso modules 104B. The combination may be attached to a selected one of base modules 102 , resulting in skeletal frame 130 .

The elements may be fastened together in conventional manner, eg by screws, bolts or the like. Spacer accessories can be placed between the modules to place the modules where desired. Once assembled, the outer skin 132 may be applied for aesthetic purposes. Outer skin 132 may be made of any suitable material, such as plastic. It should be understood that in other embodiments, the robot may be built from modules with an outer skin, and thus need not add a separate outer skin.

As shown, an exemplary robot may have modules shaped into a head module, a torso module, and a base module, although any number and configuration of modules may be possible in accordance with the present invention. It should be understood that the die set may have a shape depending on the position or function of the die set, or simulate a non-robotic design, such as an animal or humanoid. Each module can perform any specified function at any specified position and in any specified direction within the robot. This may differ from standard robots, where functional modules or accessories may only be located in one location on the robot. In such a robot, changing its position or function may require an entirely new robot design, including reprogramming software and firmware.

When the functional module 104 is selected for an intelligent robot, it may include an intelligent connector.

Once the functional module is located and coupled to the robot, the smart connector can be placed in electronic communication with the smart connector or hub on the base module 102 . This may allow operational information of the functional module 104 to be provided to the base module 102, thereby allowing the base to control the functional module in concert with the operation of the robot.

Electronic communications may be established in any known manner. Wires or cables may run internally along the skeletal frame to connect the two smart connectors. In some embodiments, such electronic communication may also be accomplished wirelessly, such as using Bluetooth or Near Field Communication (NFC).

Figure 4 shows some possible modules that may be placed on the robot, and a base unit with a base housing 120 and a base module 102 located therein. For example, a drug dispenser or a gripper arm (shown at the top of the figure). For purposes of illustration only and not limitation, other examples of possible mods and their associated accessories may include: medical equipment or appliances, sanitary equipment or appliances, transportation equipment or appliances, cooking equipment or appliances, entertainment equipment or appliances, Personal health and assistive devices or devices, or any device or device that would benefit from attachment to an intelligent robot.

Referring again to FIG. 1 , in the base module 102 there may be one or more smart Lync hubs or smart connectors. Any number and type of functional modules 104 may be connected to the base module 102 or modules. Several different possible accessories are also shown, such as trays and keyboards. Even modules created by third parties can be connected, as long as they include smart connectors with relevant operational information about the functional modules 104 .

Another basic configuration may have the base module 102 in the lowermost module. The head mod and torso mod contain designated functions. Other modules and accessories can be attached if desired.

A robot can have a base mod, and can have one or more additional "function" mods. A base module may contain (as a single unit or collectively) the following elements: • the robot's firmware, • The robot's main software, including the operating system (OS), and • "Repository" (repository of all/most needed firmware/features/functions for all/most mods).

When the base module 102 is assembled and selected functional modules are connected via the smart connector 108, the robot can use the information obtained from the smart connector to know some or all of the following: • what mod is attached, • how the modules are attached, • where it is attached, • module location, • module orientation, • What prescribed functionality is provided by the module, • Which purchased/prepaid optional features the mod can access, • What design material, if any, is associated with the mod (eg, 1st party, 2nd party, 3rd party developer, or company logo of the mod, or custom designs/logos/themes added to the mod). • Verification that the mod is licensed and not pirated or unlicensed.

Smart connector 108 can transmit this information to one or more base modules 102 . It in turn can obtain relevant information from the database and respond appropriately to allow the firmware and operating system of one or more base modules and their associated functional modules 104 to function normally.

As mentioned above, there are many options for base modules and function modules. Some mods may have similar purposes to each other, while some mods may have different purposes from each other. In an exemplary embodiment of a humanoid robot, the functional modules may be shaped into: a head, a head extension, a torso (decomposable into an upper torso, a middle torso, and a lower torso) and a base module that may house one or more The bottom unit of the group. Each of these mods can have one design or many designs to choose from. Mods can also have different orientations to choose from (for example, the head can be placed in landscape or portrait orientation, or the torso has connectors on the left and right or front and back, etc.).

Smart connectors can be coupled to or detached from physical connectors to keep the modules together. In some embodiments, the smart connector can be coupled to or detached from the power connector to provide power to the module (where applicable). As mentioned above, smart connectors can store digital data. This digital data may be transmitted by any of the following methods: • Wireless means such as Near Field Communication (NFC), Bluetooth or any known or proprietary wireless communication standard known and used, or • Wired means, such as a proprietary plug, or a known or used generic plug (example: USB).

When purchasing a modular robot, the purchaser may need to purchase the specified robot, or may need to commission the robot, or the manufacturer may contact a company or government agency to obtain a commission for the robot. In turn, the desired robot can be created, which includes the selection of modules and features that suit the specified needs/needs/desires/budget. The robot can be built like a Lego® creation, with specified/desired functions and characteristics. The selected robots are ready and working in a short time frame compared to current practices and technologies.

Examples of short time frames could be days, hours, and minutes. Another time- and cost-saving measure will be the manufacture of intelligent robots. Instead of having to build and develop each robot as a whole, only the individual modules themselves reduce manufacturing time and cost. This will also allow decentralizing the production of robots, as "Module A" can be manufactured in "Facility X" located in "Country J", while "Module B" can be manufactured simultaneously in "Facility Y" located in "Country I" Made in.

Technical execution can be accomplished in a number of ways known in the industry. According to the invention, the point is that when connecting the functional module directly or indirectly to the base module, the data stored on the functional module can be sent to the base module carrying the information required for the correct operation of the functional module. Thereafter, the base module can use this information and configure itself to operate correctly. This allows for customizable plug-and-play robots that can be built quickly.

Reference is now made to the flowchart in FIG. 5 .

A general example of data flow during the assembly and start-up phases of robot production is given below and is provided for illustrative purposes only and is not intended to be limiting.

Step 501: Data can be preloaded onto the functional module 104, for example, preloaded into the chip or part of the chip, solid state drive (SSD), hard drive (HDD), QR code, near field communication (NFC) Chips, barcodes, etc. It will be appreciated that this process may preferably be done at the time the mod is created, so it may take effect some time before the mod is actually installed on the robot. This material can be raw, or configured and encoded. The data itself can be stored in many standard ways in the field, such as: • A [eg, binary] code for each piece of information, where the number represents a specific predefined component. For example, there may be 27 different mods, each with a unique identification number (0000,0001, 0010,0011, etc.). Another [binary] code may represent a specific predefined function. For example, there may be 58 different functions, each with a unique identification number (0000,0001, 0010,0011, etc.), and so on. • A [eg, binary] code for each piece of information, where the number represents a specific predefined component or function when segmented. For example, there may be 127 different modules, and each has a unique identification number (1001-0010-1111-0100), where the first set of digits can represent a unique identification code for a set of information, such as manufacturer, part number , component function, version number, production year, warranty information, etc. • A larger set of codes, which can contain details about components and functions, does not require a predefined number representing a specific component or function. • A set of data that, knowing the database, holds information about where the data can be located in the database that may be important to the functional usage of the module and the details of the module's components. Think of this implementation as a pointer in C programming (a pointer can be a variable whose value can be the address of another variable, ie, the direct address of a memory location). As in a set of data representing a memory address or location in a database to look up for an associated module.

Step 502: Establish a connection. There may be a physical connection (could be magnets, clips, screws, sockets, links, snaps into place, anything that physically connects the two modules) and an electrical connection (could be a plug-in cable (e.g. USB), wireless, bluetooth, proximity The location to make the NFC signal readable, the QR code to be scanned, the connection of two metal dots forming a continuous wire).

Step 503: In embodiments where the Smart Connector or Smart Link 108 requires connection data (data about where and how the functional module can be attached (see below) or any other data required here), • What is attached, (sometimes known in advance) • How it is attached, (options: may be known in advance and coded in or generated at connection time) • where it is attached to, (options: may be known in advance and coded or generated at link time) • What prescribed functionality is provided by the module, (usually known in advance) • Which purchased/prepaid optional features the mod will have access to, (usually known in advance) • What design material, if any, is associated with the mod (eg, 1st party, 2nd party, 3rd party developer's company logo, or customer design/logo/theme added to the mod) (sometimes previously Know) • Verification that the mod is licensed and not pirated or unlicensed (often known/established in advance)

Data can be generated in this step. Note that in other embodiments, this material may not exist or be generated in advance once configuration instructions are given, manually or digitally added to the module and entered into step 501 . In this embodiment of generating link profiles, link profiles can be done in several different ways. For example, a connection list, or one or more base modules with designated ports representing a particular placement on the robot.

In the example of a connection list, the example executes as follows. The option to generate information about the connection can be done by each port of the smart connector with a unique address in the robot. This information can be stored if the module is connected from port to port. As shown in Figure 6, this can be a very simple way of generating information.

Through the connection list, information is generated about how and where each module can be connected for smart connection. (This example has a single base module 102 for clarity, but broader examples exist).

For example, as shown in Figure 6, Connection list: Module A path: A1->AA1 Module B path: B1->A2->A1->AA1 Module C path: C1->B2->B1->A2->A1->AA1 Module D path: D1->A3->A1->AA1 Module E path: E1->A4->A1->AA1 Module F path: F1->B4->B1->A2->A1->AA1

In the example of specifying a port, an example implementation may be as follows with reference to FIGS. 7 and 8 .

Specific Ports: The option to generate information about connections can be done for each port representing a specific placement of the base module. Thus, when a connection is established with one or more base modules, it immediately knows how it is connected. This is illustrated in Figures 7 and 8, where Figure 7 may represent the electrical connection separate from the physical connection directly into the port (e.g., wired or wireless), while Figure 8 may represent the electrical connection accompanying the physical connection (e.g., The wires run through the middle module or connect via non-smart ports (ports that just extend the wires through without any intelligence). (This example has a single base module 102 for clarity, but broader examples exist).

Step 504: The functional module 104 may be directly connected to one or more base modules 102 or indirectly connected. In the case of specific ports, this may not be a problem, since all connections to one or more dock modules are directly connected in some way.

Step 505: In the case of a connection list, the robot can have multiple indirectly connected modules, and each module can have multiple ports. Indirect connections may exist and internal data must be transferred from module port to module port before it reaches one or more base modules 102 . This can be done in a variety of ways known in the industry, by any wired or wireless means. For example, the data can be stored like a linked list (linear/non-linear linked data structure) in C programming (a linked list is a very commonly used linear data structure that consists of a set of nodes in a sequence. Each node keeps its own data and the address of the next node, thus forming a chain structure). This could be an extremely simple and non-data-intensive way of storing connection data.

Step 506: The material may eventually be received by one or more of the base modules 102; and, if it is encoded, it may be decoded (encoding-based method) into executable material in this step. Note that encoding and decoding can have broad terms here. The encoding here can be as simple as the data stored in step 501, and the decoding can be to interpret the data into usable information.

Step 507: Once the data is executable, the base module 102 can process the data and can download it from a database (by internal means, such as downloaded data on an SSD or HHD or similar) or external means (such as an external drive plugged in, Or find to configure the robot via Internet Wi-Fi, cloud, ethernet plug). One or more base modules 102 can flag important data and can additionally create/store a subset of data that the robot needs for quick lookup or access to operate the robot based on ID/configuration/connection data.

Step 508: (this step is not shown in the flowchart of FIG. 5 ). In cases where a module includes authentication material, one or more base modules 102 may verify that the module is authorized, or not unauthorized, unlicensed, or pirated. After this, the robot can be allowed to work. Note that this step can occur during step 506, after step 506 and before step 507, during step 507, and after step 507 without loss of functionality.

It should be understood that smart connectors and smart hubs can function in both non-initialized and post-initialized features and operations. The smart connector may also include a smart connector chip that holds the information discussed in step 501 . As mentioned above, the initial specification details are similar to those in Smart Connector initialization, but post-initialization features will be added here for long-term usability and example purposes.

Smart connector chips can be used, for example, to facilitate data transfer between tool modules, initiate connection of functional modules, verify the authenticity of functional modules, verify the authenticity of functions and features of functional modules, or any other information.

In the example of a smart connector chip containing authentication data, the chip can also be used to periodically verify that the robot or feature or feature of the feature is licensed, appropriate, or paid for. This can be done by requiring internet access at the beginning or end of a session to check if the bot is licensed, appropriate or paid for. Examples of time periods may include a day, a week, a month, a year, or any suitable time interval. This might be similar to how online subscriptions work for offline products. Functionality works fine offline, but requires the user to connect to the internet during certain periods of time and verify that the subscription is still valid, so users cannot pay for a single monthly subscription and keep it indefinitely.

In addition, the smart connector chip can provide maintenance or development personnel with diagnostic data, such as power usage, data usage, usage frequency, and time of highest usage. The chip could also have indicator lights for maintenance, similar to a car's dashboard, signaling data connection errors, low power transfers and other diagnostic issues. For example, if the amount of power passing through the connector falls below a threshold, a light on the die will blink, indicating a possible connection error. Similarly, if the data signal is not properly routed through the connector (ie, the signal may be interrupted or distorted for some reason), the light can blink.

In preferred embodiments, means known in the industry for autonomous locomotion may be included. Such a modular platform for robotics can be a 3-layer system consisting of hardware, software, and artificial intelligence (AI).

The hardware layer can include the robot's skeletal system and firmware, and a real-time operating system (RTOS), which can be the low-level hardware code that controls and directs the physical device.

The software (SW) platform may include a robot operating platform (ROPS). This can be the robot base operating system, and in a preferred embodiment, it can be a proprietary operating system based on the open source ROS framework. There may also be a Brain and Behavior system (B&B) that acts as a "bridge" between the RTOS and the ROPS and connects directly to the ACB elements. In a preferred embodiment, B&B may be a proprietary intelligent system for controlling and directing physical robotic devices. AI includes ACB system.

One aspect of the invention is hardware modularization. There are multiple modules for performing robot operations, and these modules can be interchangeable and detachable.

To meet the needs of the cost-sensitive and broadly diverse social robotics market, applicants have created proprietary technology that achieves the ROP foundation needed to develop custom social robots for any need at a fraction of the time and cost of conventionally designed robots Typically 80% redeployment of the "4-pillar" architecture.

Pillar 1 is the hardware architecture. This includes a configurable proprietary CPU, MCU and power board as well as a library of ROS commands.

Next comes the physical form—modular interchangeable racks.

The third pillar is the adaptive AI [artificial intelligence] engine.

Software is the fourth pillar - a universal app that can be customized for Android, IOS, Windows and MAC OS to remotely control all robot functions.

Depending on the robot's capabilities and deployment requirements, as the central processor, the platform can support multiple CPU kits, ranging from a Raspberry Pi Compute Die 3 for more basic single-tasking applications, to the state-of-the-art Nvidia Jetson Nano or Nvidia Jetson Xavier NXAI supercomputer CPU module to provide advanced dynamic neural network (DNN) integration.

Such a modular robot can support a wide variety of MCUs with different performance and characteristics from major manufacturers, including but not limited to Microchip Atmel 8-bit AVR RISC-based microcontrollers, NXP ARM Cortex 32-bit microcontrollers, Cypress 32-bit Arm Cortex Cortex-M0+PSoC.

The system can support all currently available commercial motors. The supported protocols are: I2C, SPI, RS232, RS485, CanBus, TCP/IP, 1-wire. Mods for other protocols can be easily integrated.

Such a modular architecture can be designed to be fully scalable, which could make any service robot fully upgradeable, eliminating obsolescence.

It currently supports: - Multiple CPUs that can be used as different parts of the platform or as a cluster computer to multiply system performance. - Modular power supply unit provides virtually unlimited power channels. - Unlimited number of different types of drives.

Different software packages developed by applicants or third parties can be integrated to extend the capabilities of any service robot. In addition to in-house developed modules for natural language processing, image processing, robot navigation packages, etc., the modular platform can support almost any major AI platform (Google, Amazon, IBM, Microsoft, Apple, and many other AI platform), multiple different third-party online and offline services, sensors, switches, devices, etc.

Bots use many different processes or applications running in multi-threaded processes. This may require developing multiple process "applications" for each sensor\function and synchronizing all "tasks" without running into conflicts, which may require significant CPU processing and memory management.

The base module 102 can monitor and run all processes at all times. The AI can then decide what "behavior" may be required based on the perceptual data it receives as a single "task". Each behavior can store the correct set of steps to be executed by the system, run completely independently and simultaneously monitored by the base module. This architecture could be much more efficient, allowing the robot to dynamically support multiple behaviors and adapt to new ones.

The architecture can provide the flexibility to write "behaviors" - high-level programming (which can be done via a drag-and-drop method) instead of low-level code but with full access to low-level code when needed. Writing new behaviors can be very simple and fast. Such an approach may be relatively easy to teach compared to traditional approaches, which may support the use of open source APIs that enable any third party to define and enforce custom behavior.

One or more base modules 102 may be composed of 3 distinct layers: environment sensing layer, behavioral and self-regulating cognitive behavioral layer. In general, the first layer in one or more base modules 102 may be the ability to sense the environment in which the robot operates. This goal can be achieved by using a variety of environmental sensors: • Microphone array – for commands and voice recognition; • 12MP camera and 3D camera – for facial and object recognition, and other image and video processing needs; • Emergency sensors, such as smoke and gas detectors; and/or • LiDAR sensors, cliff sensors, infrared sensors for navigation and collision avoidance. Each sensor can be upgraded for specific customer needs.

A second layer in one or more base modules 102 may be the ability to define different behaviors performed by the robot. Behaviors are the way to simplify every task needed from a system into a set of state machine type commands (based on FlexBE2). Each behavior can use any available sensor to complete the basic tasks that the robot can perform. Each state in the state machine can contain nodes - which can essentially be threads of execution running specific compilers/scripts written in any language supported by the OS (such as C++, Python and Java) to implement the different nodes.

According to the present invention, a third layer in one or more base modules 102 may use machine learning and deep learning algorithms to better understand the robot's surroundings and assist the operator in its work. Sensing capabilities may be present. Deep learning algorithms for object recognition and facial recognition have been implemented on many different devices.

Object Recognition - Gradually mature YOLOv43/Faster R-CNN4 and tested algorithms can exist in the modular robot disclosed in this paper. Although the two algorithms are based on similarities—using an anchor-based network structure and bounding box regression—Faster R-CNN generally fails for on-the-fly detection, while YOLO has problems detecting smaller objects. The modular robot in this paper is able to detect 50 different objects with high confidence, which can be the daily necessities of residents.

For facial recognition - the modular robot in this article can use the DLIB library, which extracts 128 feature vectors, each point representing a point of interest on the object's face. Using K-NN, which calculates how similar a new face the robot sees to faces it already knows. The feature extraction process can be done using HOG or Resnet models - the main difference is the number of features to be extracted.

Fall detection - This article's modular robot uses built-in sensors to detect falls - primarily a camera.

It will be apparent to those skilled in the art that the invention is not limited to details of the foregoing illustrative embodiments, but that the invention may be embodied in other specific forms without departing from its spirit or essential attributes. Therefore, the present embodiments should be considered in all respects as illustrative rather than restrictive, the scope of the present invention is indicated by the appended claims rather than by the foregoing description, and therefore within the equivalent meaning and range of the claims All changes to are intended to be included therein.

102: Base module 104: Functional module 104A: Head module 104B: Torso Module 106: Attachments 108:Smart Connector 110:Smart hub 120: Base shell 130: Skeleton frame 132: outer skin 501, 502, 503, 504, 505, 506, 507: steps A1, A2, A3, A4, B1, B2, B3, B4, C1, D1, E1, E2, F1, AA1, AA2, AA3, AA4, AA5, AA6, AA7, AA8: ports

In order to better understand the invention and demonstrate how it can be practiced, reference will now be made to the accompanying drawings, by way of example only, in which it is emphasized that the particulars shown are shown by way of example and for purposes of illustration only of preferred embodiments of the invention. It is presented for purposes of discussion and to provide what is believed to be the most useful and understandable description of the principles and conceptual aspects of the invention.

FIG. 1 is a diagram showing the architecture of a modular robot.

Fig. 2 is a diagram showing the electrical connection of elements of the modular robot.

Fig. 3 is a view showing a skeletal frame of the modular robot.

Fig. 4 is a diagram showing different functional modules that can be added to a modular robot.

Figure 5 is a flow chart of a process showing how the invention works.

Figure 6 is a diagram illustrating how modules may be connected.

7 and 8 are connection diagrams showing how the module is connected to the base module.

102: Base module

104: Functional module

106: Attachments

110:Smart hub

Claims (17)

A modular framework for intelligent robots, including: a base that controls the motion and functions of the robot and contains memory for operating instructions for a plurality of modules, each module performing a unique function, and having smart connectors; One or more devices or modules for performing specified functions and having an intelligent connector containing a unique code for the device or module and firmware for the operation of the module ;and wherein when a module is coupled to the modular frame, the smart connector of the module communicates electronically with the smart connector of the base, thereby providing sufficient operational information for the base to operate the intelligent robot. The modular framework as claimed in claim 1, wherein the base includes a robot basic operating system and hardware codes, and the robot basic operating system and hardware code control and command the modules connected to the robot The physical devices and modules of the framework. The modular framework as described in claim 1, wherein the base includes a CPU, an MCU, a power board, a robot operating system command library, and an artificial intelligence engine. The modular framework as claimed in claim 1, wherein the base further includes a communication protocol. The modular framework of claim 1, further comprising means for autonomous movement. A modular kit for generating an intelligent robot, the modular kit comprising: A number of optional base modules including: robot firmware, which includes low-level hardware code that controls and directs at least said base module; a robot operating system (OS), which includes a plurality of operating instructions for a plurality of functional modules; a database containing firmware, features and functions of the base module and the function module; and at least one of a smart Lync hub and a smart connector capable of connecting to multiple smart connectors simultaneously; and A plurality of selectable function modules, which are used to perform specified functions, each selectable function module has a smart connector, and the smart connector contains a unique code for the function module and is used for the The firmware for the operation of optional feature modules. As the kit described in request item 6, the plurality of base modules also include smart firmware, communication protocol, CPU, MCU, power board, robot operating system command library, artificial intelligence engine, memory and communication module at least one of the . The kit according to claim 6, wherein the functional module operates at any specified position in the intelligent robot. The kit according to claim 6, said functional modules operate in any specified direction. In the kit according to claim 6, the functional module communicates electronically with the base module via the smart connector. The kit according to claim 6, wherein the functional module is a drug dispenser, a clamping arm, a torso element, a medical device or device, a sanitary device or device, a transportation device or device, a cooking device or device, an entertainment device or Appliances and at least one of personal health and assistive devices. The kit as claimed in claim 6, wherein the smart connector includes a unit for providing sufficient operation information to the base module to enable the kit to operate as an intelligent robot. The kit of claim 6, wherein the smart connector includes at least one of a wireless communication unit and a wired connection. The kit of claim 13, wherein the wired connection is at least one of a direct connection, a proprietary connector, and a non-proprietary connector. The kit according to claim 13, wherein the smart connector includes a unit for transmitting an attached module identification, an attached module attachment method, an attached module location, an attached module At least one of the specified function of the attached module, the optional function of purchasing/prepaying the attached module, the associated design information of the attached module, and the attached module verification information. The kit of claim 15, wherein the associated design material is at least one of a logo, a theme, and a design. The kit of claim 15, wherein the authentication material is at least one of licensed, unlicensed, and pirated.

Images