Abstract: A computer-implemented method, system and computer program product for refactoring code using a machine learning model. Parallel corpora is generated using a single directional code transform. A single directional code transform refers to a transformation performed by a refactoring tool which refactors computer code (“code”) to restructure the code to include styles, which are often undesirable, such as “for” loops. Parallel corpora refers to a collection of code of a first style of code (e.g., dictionary comprehensions) in the code prior to refactoring (non-refactored code) and a second style of code (e.g., “for” loops) in the refactored code. A machine learning model is then trained to perform code refactoring in a reverse direction of the single directional code transform using the parallel corpora. New computer code is then refactored using the trained machine learning model, where the refactored code now includes a desired style of code (e.g., dictionary comprehensions).

Abstract: A method for monitoring an industrial robot. The method includes configuring the robot to perform a certain task during an integration process and storing integration data in the robot identifying the configuration of the robot for performing the task. The method also includes installing the robot in a manufacturing facility, and uploading the stored integration data to the Cloud when the robot is installed in the manufacturing facility. The method further includes capturing production data generated by the robot during operation of the robot in the manufacturing facility, uploading the production data to the Cloud, and comparing the production data to the integration data.

Abstract: Techniques regarding transferring learning outcomes across machine learning tasks in automated machine learning systems are provided. For example, one or more embodiments described herein can comprise a system, which can comprise a memory that can store computer executable components. The system can also comprise a processor, operably coupled to the memory, and that can execute the computer executable components stored in the memory. The computer executable components can comprise a transfer learning component that can executes a machine learning task using an existing artificial intelligence model on a sample dataset based on a similarity between the sample dataset and a historical dataset. The existing artificial intelligence model can be generated by automated machine learning and trained on the historical dataset.

Abstract: In an approach to improve detecting and correcting errors in one or more machine learning pipelines. Embodiments comprise generating a plurality of test machine learning pipeline instances based upon a target machine learning pipeline and evaluating the plurality of test machine learning pipeline instances for failure in a task. Further, embodiments identify one or more root causes of error based upon the evaluated plurality of test machine learning pipeline instances and failure in the task, and create a remediated target machine learning pipeline based upon the identified one or more root causes of error. Additionally, embodiments output the remediated machine learning pipelines.

Abstract: Systems, computer-implemented methods, and computer program products to facilitate inter-operator backpropagation in AutoML frameworks are provided. According to an embodiment, a system can comprise a processor that executes computer executable components stored in memory. The computer executable components comprise a selection component that selects a subset of deep learning and non-deep learning operators. The computer executable components further comprise a training component which trains the subset of deep learning and non-deep learning operators, wherein deep learning operators in the subset of deep learning and non-deep learning operators are trained using backpropagation across at least two deep learning operators of the subset of deep learning and non-deep learning operators.

Abstract: A machine-learning model task deduction method, system, and computer program product include extracting data schema of a machine-learning model and analyzing the data schema to determine an intended task of the machine-learning model.

Abstract: In an approach to improve detecting and correcting errors in one or more machine learning pipelines. Embodiments comprise generating a plurality of test machine learning pipeline instances based upon a target machine learning pipeline and evaluating the plurality of test machine learning pipeline instances for failure in a task. Further, embodiments identify one or more root causes of error based upon the evaluated plurality of test machine learning pipeline instances and failure in the task, and create a remediated target machine learning pipeline based upon the identified one or more root causes of error. Additionally, embodiments output the remediated machine learning pipelines.

Abstract: An augmented reality (AR) system for production-tuning of parameters for a visual tracking robotic picking system. The robotic picking system includes one or more robots configured to pick randomly-placed and randomly-oriented parts off a conveyor belt and place the parts in an available position, either on a second moving conveyor belt or on a stationary device such as a pallet. A visual tracking system identifies position and orientation of the parts on the feed conveyor. The AR system allows picking system tuning parameters including upstream, discard and downstream boundary locations to be visualized and controlled, real-time robot pick/place operations to be viewed with virtual boundaries, and system performance parameters such as part throughput rate and part allocation by robot to be viewed. The AR system also allows virtual parts to be used in simulations, either instead of or in addition to real parts.

Abstract: Embodiments are disclosed for a method. The method includes identifying one or more source code signals in a source code. The method also include generating an amplified code based on the identified signals and the source code. The amplified code is functionally equivalent to the source code. Further, the amplified code includes one or more amplified signals. The method additionally includes providing the amplified code for a machine learning model that is trained to perform a source code relevant task.

Abstract: A method and system for programming a path-following robot to perform an operation along a continuous path while accounting for process equipment characteristics. The method eliminates the use of manual teaching cycles. In one example, a dispensing robot is programmed to apply a consistent bead of material, such as adhesive or sealant, along the continuous path. A computer-generated definition of the path, along with a model of dispensing equipment characteristics, are provided to an optimization routine. The optimization routine iteratively calculates robot tool center point path and velocity, and material flow, until an optimum solution is found. The optimized robot motion and dispensing equipment commands are then provided to an augmented reality (AR) system which allows a user to visualize and adjust the operation while viewing an AR simulation of dispensing system actions and a simulated material bead. Other examples include robotic welding or cutting along a continuous path.

Abstract: An augmented reality (AR) system for diagnosis, troubleshooting and repair of industrial robots. The disclosed diagnosis guide system communicates with a controller of an industrial robot and collects data from the robot controller, including a trouble code identifying a problem with the robot. The system then identifies an appropriate diagnosis decision tree based on the collected data, and provides an interactive step-by-step troubleshooting guide to a user on an AR-capable mobile device, including augmented reality for depicting actions to be taken during testing and component replacement. The system includes data collector, tree generator and guide generator modules, and builds the decision tree and the diagnosis guide using a stored library of diagnosis trees, decisions and diagnosis steps, along with the associated AR data.

Abstract: A method for automatically generating documentation for an artificial intelligence model includes receiving, by a computing device, an artificial intelligence model. The computing device accesses a model facts policy that indicates data to be collected for artificial intelligence models. The computing device collects artificial intelligence model facts regarding the artificial intelligence model according to the model facts policy. The computing device accesses a factsheet template. The factsheet template provides a schema for an artificial intelligence model factsheet for the artificial intelligence model. The computing device populates the artificial intelligence model factsheet using the factsheet template with the artificial intelligence model facts related to the artificial intelligence model.

Abstract: Techniques regarding transferring learning outcomes across machine learning tasks in automated machine learning systems are provided. For example, one or more embodiments described herein can comprise a system, which can comprise a memory that can store computer executable components. The system can also comprise a processor, operably coupled to the memory, and that can execute the computer executable components stored in the memory. The computer executable components can comprise a transfer learning component that can executes a machine learning task using an existing artificial intelligence model on a sample dataset based on a similarity between the sample dataset and a historical dataset. The existing artificial intelligence model can be generated by automated machine learning and trained on the historical dataset.

Abstract: A machine-learning model task deduction method, system, and computer program product include extracting data schema of a machine-learning model and analyzing the data schema to determine an intended task of the machine-learning model.

Abstract: A method for automatically generating documentation for an artificial intelligence model includes receiving, by a computing device, an artificial intelligence model. The computing device accesses a model facts policy that indicates data to be collected for artificial intelligence models. The computing device collects artificial intelligence model facts regarding the artificial intelligence model according to the model facts policy. The computing device accesses a factsheet template. The factsheet template provides a schema for an artificial intelligence model factsheet for the artificial intelligence model. The computing device populates the artificial intelligence model factsheet using the factsheet template with the artificial intelligence model facts related to the artificial intelligence model.

Abstract: A robot is moved along a first continuous programmed path with a robot controller executing a learning path control program without performing an operation on a workpiece. The actual movement of the robot along the first continuous programmed path is recorded. The first continuous programmed path is adjusted to create a second programmed path. The robot is moved along the second continuous programmed with the robot controller executing the learning path control program without performing the operation on the workpiece. The actual movement of the robot along the second continuous programmed path is recorded. Traces of the recorded actual movements of the robot along the first continuous programmed path and the second continuous programmed path are displayed.

Abstract: A method and a system stream robot tool center point position to external processors at high frequency. The method includes the steps of: reading robot joint encoder data using an Interrupt Service Routine in the robot controller; calculating tool center point position based on the encoder data; and sending the calculated position data to a network socket in a high priority task. The method achieves tool center point and/or joint position communication at fast and consistent time intervals, as compared to much longer times for prior art methods. A downstream device, such as a processor or controller for another machine, reads the communicated tool center point and/or joint position data and uses it to control the operations of its own device. High speed motion command streaming from outside processors can be used in a similar way to control the robot.

Abstract: A method of deploying artificial intelligence (AI) model resources includes storing at least one AI model in a model store memory in a plurality of different versions, each different version having a different level of fidelity. When a request to exercise the AI model is received, a processor determines which version of the AI model to exercise for the received request. The determined AI model version is used to serve the received request by exercising input data accompanying the received request. The result of the exercised AI model version is used to respond to the received request.

Abstract: Methods and systems for touch-sensing to provide an updated user frame are provided. These include the provision of a user frame and the touch-sensing of a workpiece, where the touch-sensing includes performing a touch-sensing schedule. The touch-sensing schedule includes one of a laser touch-sensing event and a wire touch-sensing event, where one of the laser touch-sensing event and the wire touch-sensing event is switched to the other of the laser touch-sensing event and the wire touch-sensing event while performing the touch-sensing schedule. An offset of the workpiece relative to the user frame is determined based on the touch-sensing of the workpiece and the offset is applied to the user frame to provide the updated user frame. The unique dynamic user frame feature enables same touch sensing program to be cloned and applied on multiple robot controllers.

Abstract: A robot data transfer method includes the step of collecting data from each of a plurality of robots in a multi-robot production facility in real-time. The data collected from the robots is then transferred in real-time from a controller of each of the robots to a first data collection device. Within the first data collection device, the data is buffered using a multi-segment queueing mechanism. The queueing mechanism is configured with a retention policy. The data is then transferred to a second data collection device based on the retention policy of the queueing mechanism of the first data collection device. The second data collection device analyzes the data and determines whether maintenance or optimization is necessary for any of the robots.