Abstract: A powertrain system may determine a power distribution for a set of power sources of a vehicle. The powertrain system may be coupled to a perception system that may provide perception data indicating a scenario, situation, or environment that has been encountered by the vehicle. The powertrain system may also receive health values. The powertrain system may include machine learning model that may generate the power distribution based on one or more of the perception data, the health values, and a power request.

Abstract: Methods and systems for detecting defect in sample cartridges in real-time during manufacturing. Such systems utilize one or more external sensors that detect characteristics or parameters of the sample cartridge and/or the manufacturing process from one or more data sets. The external sensor(s) include any of: an RGB camera, IR camera, high-resolution optical camera, and ultrasonic microphone or combination thereof. An automated system obtains data sets from external sensor(s) and compares the data sets to a baseline of the sample cartridge and/or manufacturing process such that defects can be determined based on a variance from the baseline. Such methods can utilize feature extraction and spectrum analysis to identify features or characteristics for comparison with the baseline. A machine learning model can be used to determine an algorithm based on data sets of acceptable sample cartridges and data sets from the external sensor(s) that are associated with the cartridge defect.

Abstract: Methods and systems for training a model for automated defect detection of a product during manufacturing are provided herein. Such methods utilize a combination of supervised transfer learning through auxiliary tasks and a combination of supervised and unsupervised learning. The methods can utilize supervised transfer learning with expert labels on a generalized auxiliary task, such as product classification, which is transferred to more specific auxiliary tasks, such as identification of specific product features and/or anomaly detection, where additional expert labels are then applied to the anomalies, and another iteration of supervised learning further improves the model. The anomalies can correspond to features associated with defects, which can be induced experimentally to improve efficiency of the training procedure. The product can be a sample cartridge such that the model allows detection of faulty cartridges based on sample cartridge and/or manufacturing process data.

Abstract: Disclosed embodiments provide a technical improvement for providing localization for a transportation vehicle by detecting road wear reference lines in a roadway on which the transportation vehicle is travelling and controlling, guiding or otherwise facilitating alignment of the transportation vehicle wheel centers with the detected centers of the road wear.

Abstract: In one embodiment, a method is provided. The method includes obtaining a sequence of images of a three-dimensional volume of a material. The method also includes determining a set of features based on the sequence of images and a first neural network. The set of features indicate microstructure features of the material. The method further includes determining a set of material properties of the three-dimensional volume of the material based on the set of features and a first transformer network.

Abstract: A system and a method are provided for compressing a deep neural network (“DNN”). In some examples, the DNN is trained, where the DNN has at least one layer having multiple filters. Clustering of the filters of at least one layer is performed. Dimension reduction can be applied as well to the filters to reduce the channel dimensionality of the at least one layer. The dimensionally reduced DNN can then be retrained. Once retrained, the compressed DNN can be stored in a storage device.

Abstract: A method, apparatus, system for batter material screening is disclosed. First, microstructure generation parameters for a plurality of microstructures are received, where the microstructure generation parameters include microstructure characteristics. Microstructure statistics are generated using a first artificial intelligence (“AI”) model, where the received microstructure generation parameters are inputs for the first AI model. Microstructure properties are predicted using a second AI model for the microstructures based on the generated microstructure statistics, the received microstructure generation parameters, and battery cell characteristics. It is determined whether at least one of the microstructures is within a predefined energy profile range based on the predicted microstructure properties.

Abstract: A powertrain system may determine a power distribution for a set of power sources of a vehicle. The powertrain system may be coupled to a perception system that may provide perception data indicating a scenario, situation, or environment that has been encountered by the vehicle. The powertrain system may also receive health values. The powertrain system may include machine learning model that may generate the power distribution based on one or more of the perception data, the health values, and a power request.

Abstract: A computing system, method, and apparatus for determining a state of health indication for a battery are provided. A first supervised deep neural network (“DNN”) is trained using received characteristics for the battery as input and received process parameters as outputs. An unsupervised AI estimator is trained using one or more clustering methods based on extracted features from the first supervised DNN, where the received characteristics are input to the unsupervised AI estimator. A second supervised DNN is trained using identified clusters from the unsupervised AI estimator. The identified clusters are validated with state of health indications. User battery data is inputted to the second supervised DNN to determine the state of health for the battery.

Abstract: A method, apparatus, system for batter material screening is disclosed. First, microstructure generation parameters for a plurality of microstructures are received, where the microstructure generation parameters include microstructure characteristics. Microstructure statistics are generated using a first artificial intelligence (“AI”) model, where the received microstructure generation parameters are inputs for the first AI model. Microstructure properties are predicted using a second AI model for the microstructures based on the generated microstructure statistics, the received microstructure generation parameters, and battery cell characteristics. It is determined whether at least one of the microstructures is within a predefined energy profile range based on the predicted microstructure properties.

Abstract: A computing system, method, and apparatus for determining a state of health indication for a battery are provided. A first supervised deep neural network (“DNN”) is trained using received characteristics for the battery as input and received process parameters as outputs. An unsupervised AI estimator is trained using one or more clustering methods based on extracted features from the first supervised DNN, where the received characteristics are input to the unsupervised AI estimator. A second supervised DNN is trained using identified clusters from the unsupervised AI estimator. The identified clusters are validated with state of health indications. User battery data is inputted to the second supervised DNN to determine the state of health for the battery.

Abstract: A system and a method are provided for compressing a deep neural network (“DNN”). In some examples, the DNN is trained, where the DNN has at least one layer having multiple filters. Clustering of the filters of at least one layer is performed. Dimension reduction can be applied as well to the filters to reduce the channel dimensionality of the at least one layer. The dimensionally reduced DNN can then be retrained. Once retrained, the compressed DNN can be stored in a storage device.

Abstract: Disclosed embodiments provide a technical improvement for providing localization for a transportation vehicle by detecting road wear reference lines in a roadway on which the transportation vehicle is travelling and controlling, guiding or otherwise facilitating alignment of the transportation vehicle wheel centers with the detected centers of the road wear.