Software defined manufacturing/assembly system
The present system is a software defined manufacturing (SDM) system that integrates several technologies and methods into a system that automates the process of engineering and operating automated manufacturing systems (aka “automating automation”). In one embodiment, some or all of the below aspects of the “automating automation” system are integrated: modular, configurable, reusable manufacturing cells; computer vision systems; autocalibration systems; a recipe-based programming environment; configuration management system; production analytics; and a marketplace for sharing recipes.
1. A software defined manufacturing system comprising:
a physical robotic cell including a plurality of sensors, the physical robotic cell controlled by a recipe instructing the physical robotic cell to perform a process;
a memory to collect operational data from the physical robotic cell, the operational data including the process and concurrent data from the plurality of sensors;
a processor implementing a digital twin of the physical robotic cell, the digital twin configured with sensor data and state data from the physical robotic cell to represent the physical robotic cell;
wherein the recipe is tested on the digital twin, and the recipe is then remotely deployed to the physical robotic cell and a set of dependencies is identified for a particular command or action in the recipe;
wherein the digital twin is used to validate and monitor the physical robotic cell and provide a visual representation of the physical robotic cell, the digital twin configured to enable previewing of actions of the physical robotic cell during a process, track the actions of the physical robotic cell during the process, and review the actions after the process, and provide a non-visual representation of the physical robotic cell; and
an integrated debugger to enable debugging the physical robotic cell utilizing the recipe and the operational data from the execution of the recipe of the physical robotic cell, the integrated debugger configured to display a step-by-step execution of the recipe, with associated operational data on the digital twin;
wherein data aggregation collects and logs data from cameras and the plurality of sensors in the physical robotic cell for training, tracking effectiveness of the recipe by machine learning; and
a machine learning and self-improvement system to iteratively alter settings in the physical robotic cell, monitor and evaluate results of the alteration, and provide continuous improvement of process parameters.
2. The software defined manufacturing system of claim 1 , further comprising:
a configuration management system to track a state of the physical robotic cell, the configuration management system enabling deploying changes to the physical robotic cell, and rolling back changes from the physical robotic cell.
3. The software defined manufacturing system of claim 2 , wherein the state of the physical robotic cell includes hardware, software, settings, and the recipe.
4. The software defined manufacturing system of claim 2 , further comprising:
deployment management system to enable deployment of one or more configuration changes to the physical robotic cell, wherein the deployment management system enables deployment to a selected one or more robotic cells via a cloud-based system.
5. The software defined manufacturing system of claim 1 , further comprising:
wherein the physical robotic cell is calibrated using the sensors, and calibration data is used to update the digital twin.
6. The software defined manufacturing system of claim 1 , further comprising:
development tools to create the recipe using declarative programming;
a compiler to compile the recipe to robotic commands; and
one or more drivers associated with devices of the physical robotic cell, to translate the robotic commands to device-specific commands, thereby enabling device agnostic recipes.
7. The software defined manufacturing system of claim 1 , further comprising:
a programming interface to create the recipe based on data obtained from the physical robotic cell, the programming interface providing an application user interface based on a topology of the physical robotic cell.
8. The software defined manufacturing system of claim 7 , wherein the recipe is compiled to device-specific commands, which enables use of a heterogeneous set of parts within the physical robotic cell, abstracting the recipe from a particular device definition.
9. The software defined manufacturing system of claim 1 , further comprising:
one or more physical robotic cells defining a physical micro-factory to perform a process;
wherein individual robotic cells in the physical micro-factory can be reconfigured, such that the physical micro-factory continues to function when a new robotic cell is added or a robotic cell is removed from the physical micro-factory.
10. The software defined manufacturing system of claim 1 , further comprising:
an inspection system configured to utilize the sensors in the physical robotic cell to provide in-process inspection which need not be directly connected to a current process.
11. The software defined manufacturing system of claim 1 , further comprising:
a design-for-manufacture-or-automation (DFMA) system to use develop computer aided design (CAD) data for a product to be manufactured based on manufacturing capabilities of the physical robotic cell.
12. The software defined manufacturing system of claim 11 , wherein the CAD data is refined based on the sensor data and the state data from the physical robotic cell.
13. The software defined manufacturing system of claim 1 , further comprising:
automatically generating a portion of the recipe based on data from a computer aided design (CAD) system.
14. The software defined manufacturing system of claim 1 , wherein the integrated debugger enables one or more of: remote live debugging of the physical robotic cell using the sensor data, replay of a set of steps on the digital twin using the sensor data, and simulation debugging on the digital twin.
15. The software defined manufacturing system of claim 1 , further comprising:
the integrated debugger further configured to enable stepwise debugging while controlling the physical robotic cell.
16. The software defined manufacturing system of claim 1 , further comprising:
a micro-factory enabling reconfiguration of individual robotic cells, to continue functioning when a robotic cell is added to, or removed from, the micro-factory.
17. The software defined manufacturing system of claim 1 , further comprising:
a network interface in the physical robotic cell, enabling the physical robotic cell to be upgraded remotely.
18. The software defined manufacturing system of claim 1 , further comprising:
the integrated debugger further configured to provide instant replay for debugging, enabling the digital twin to mirror each of the actions of the physical robotic cell at varying speeds based on the operational data and the recipe, to enable investigation of what occurred during a process.
19. A computer-implemented method of operating software defined manufacturing system comprising:
executing a recipe in a physical robotic cell including a plurality of sensors, the recipe instructing the physical robotic cell to perform a process;
collecting operational data from the physical robotic cell, the operational data including the process and concurrent data from the plurality of sensors;
implementing a digital twin of the physical robotic cell, the digital twin configured with sensor data and state data from the physical robotic cell to represent the physical robotic cell;
wherein the recipe is tested on the digital twin, and the recipe is then remotely deployed to the physical robotic cell and a set of dependencies is identified for a particular command or action in the recipe;
wherein the digital twin is used to validate and monitor the physical robotic cell and provide a visual representation of the physical robotic cell, the digital twin configured to enable previewing of actions of the physical robotic cell during a process, track the actions of the physical robotic cell during the process, and review the actions after the process, and provide a non-visual representation of the physical robotic cell; and
debugging, using an integrated debugger, the physical robotic cell utilizing the recipe and the operational data from the execution of the recipe of the physical robotic cell, the integrated debugger configured to display a step-by-step execution of the recipe, with associated operational data on the digital twin;
collecting and logging data from cameras and the plurality of sensors in the physical robotic cell for training, tracking effectiveness of the recipe by machine learning; and
iteratively altering settings in the physical robotic cell, monitoring and evaluating results of the alteration, and providing continuous improvement of process parameters by a machine learning and self-improvement system.
20. The computer-implemented method of claim 19 , further comprising:
combining a plurality of physical robotic cells to define a physical micro-factory to perform a second process;
wherein one or more of the physical robotic cells in the physical micro-factory can be reconfigured for executing the second process.