Abstract: Machine learning model optimization systems and methods are disclosed. A system receives sensor data captured by one or more sensors of a vehicle during a first time period. The vehicle uses a first trained machine learning (ML) model for one or more decisions of a first decision type during the first time period. The system generates a second trained ML model at least in part by using the sensor data to train the second trained ML model. The system identifies an optimal trained ML model from a plurality of trained ML models. The plurality of trained ML models includes the first trained ML model and the second trained ML model. The system causes the vehicle to use the optimal trained ML model for one or more further decisions of the first decision type during a second time period after the first time period.

Abstract: A sequence of images and a vehicle location associated with each of the images is received at a traffic light ROI management system. At least one traffic light is detected in each image. A ECS traffic light ROI is defined for each image. The ECS traffic light ROI encloses the detected traffic lights. A visual feature template is generated for each image. The visual feature template is based on the ECS traffic light ROI for the image. Each visual feature template is mapped to the vehicle location associated with the image to a HD map. The HD map is transmitted to an autonomous vehicle to enable the autonomous vehicle to identify a real-time traffic light ROI in a real-time image based on a match between a first visual feature template and real-time visual features of the real-time image at the vehicle location associated with the first visual feature template.

Abstract: A dual rate torsion bar suspension system is connected to a control arm and includes a first rate torsion bar and a second rate torsion bar connected to one another in order to provide the suspension system with a dual rate torque resistance. A cam mechanism is employed for isolating the second rate torsion bar from the first rate torsion bar.

Abstract: A device for supporting a moveable load bearing component includes a base configured to be fixedly disposed at a vehicle support structure and located proximate to the load bearing component, the load bearing component configured to be moved between a first position and a second position. The device also includes a support member connected to the base and moveable between a retracted position and a load position, the support member supporting the load bearing component when the support member is in the load position. The support member is configured to automatically move between the retracted position and the load position based on the load bearing component being moved between the first position and the second position.

Abstract: The disclosed technology provides solutions for measuring seal characteristics and in particular, for precisely measuring frictional characteristics of a rotary seal. In some aspects, a seal testing apparatus of the disclosed technology can include a motor coupled to a proximal end of a shaft, a rotator coupled to a distal end of the shaft, and a housing disposed around an exterior surface of the rotator, wherein the housing is configured to removably receive at least one seal. In some aspects the seal-testing apparatus can further include a torque transducer coupled to the shaft, wherein the torque transducer is configured to measure a torque output of the motor.

Abstract: Automatic methods and robotic system for processing workpieces are provided. A method for locally re-shaping a moldable workpiece includes receiving data in an input module indicating a present form of the workpiece; providing a robot configured to pass the workpiece through a machine; determining, with a control module, a path of the robot based on the present form of the workpiece and a target form of the workpiece; and performing a finishing process by passing, with the robot, the workpiece through the machine along the path, wherein the machine re-shapes the workpiece with the target form.

Abstract: Methods of mapping buildings for first responders, include using a smart device app to identify data related to entrances and exits in the building; storing the data in a centralized location or smart device app; and providing the data from the centralized location to the first responders in emergency situations. The methods may include recognizing patterns of the entrances and exits with the smart device app, the smart device app collects images, such that important objects recognizing in the collected images with the smart device app. A drone may be used to assist in mapping the building and measurements, such that the images collected by the drone are subject to important object recognition. The data to create augmented reality maps that are provided to first responders in emergency situations. Third-party devices may contribute to mapping the building and external factors mapped, including type of access to house, driveway, elevation.

Abstract: Compact electric drive units and vehicles with such drive units are provided. An exemplary electric drive unit includes an electric motor configured to provide an electric motor torque; an input differential configured to split the electric motor torque into a first path and a second path; a final drive gear on each path; a pinion gear on each path; and a rear transfer shaft assembly and a front transfer shaft assembly on each path. Further, each pinion gear is configured to transfer torque to a respective rear transfer shaft assembly and a respective front transfer shaft assembly, and the respective rear transfer and front transfer shaft assemblies are configured to transfer torque to the respective final drive gear.

Abstract: Presented are intelligent vehicle control systems enabling external control of vehicles using visible or audible cues, methods for making/using such vehicle control systems, and motor vehicles equipped with such control systems. A method of controlling operation of a vehicle includes a vehicle controller receiving a trigger signal indicating the vehicle is in an automatic control trigger state and, responsive to receiving the trigger signal, activating an automatic control mode that enables an entity outside the vehicle to control the vehicle using visible and/or audible cues. The vehicle controller determines if an on-vehicle network of sensing devices detects a visible/audible cue from the external entity; if so, the controller responsively determines if the detected visible/audible cue is a preset valid command.

Abstract: The single-channel, Speech Features-Based Voice Activity Detection (SFVAD) system is a robust, low-latency system that generates per-frame speech and noise indications, along with calculating a pair of speech and noise time-frequency masks. The SFVAD system controls an adaptation mechanism for a Beam-Forming system control module and improves the speech quality and noise reduction capabilities of Automatic Speech Recognition applications, such as Virtual Assistance (VA) and Hands-Free (HF) calls, by robustly handling transient noises. The system extracts speech-like patterns from an input audio signal and it is invariant to the power-level of the input audio signal. Noise calculation is controlled by a pair of speech features-based detectors (voiced and unvoiced). A Cepstral-based pitch detector and a Centrum calculation method are used to prevent contamination of the calculated noise by speech content.

Abstract: A prismatic battery cell assembly has a prismatic container that includes a body portion composed of opposed first and second side portions, opposed first and second end portions, a bottom portion, and a top portion. An electrode stack is composed as anodes interleaved with cathodes. Each of the anodes includes a first current collector having an uncoated anode current collector portion that extends in a first direction away from the electrode stack, and each of the cathodes includes a second current collector having an uncoated cathode current collector portion that extends in a second direction away from the electrode stack. A first terminal is joined to the uncoated anode current collector portions of the first current collectors of the anodes, and a terminal is joined to the uncoated cathode current collector portions of the second current collectors of the cathodes.

Abstract: A system and method for estimation of a driver state based on eye gaze includes, capturing and sending, using an outward looking camera situated in a vehicle, a first video stream of surrounding environment to a neural controller. The neural controller, based on the first video stream, generates an expected gaze distribution. Using an inward looking camera situated in the vehicle, the camera captures and sends a second video stream of a face of a driver to an eye tracker controller, where based on the second video stream, the eye tracker controller extracts a plurality of gaze directions. A gaze distribution module generates, based on the plurality of gaze directions, an actual gaze distribution. A distance distribution controller, based on a difference between the expected gaze distribution and the actual gaze distribution, generates a distance measure where a determination is made that the distance measure exceeds a threshold.

Abstract: A voiceprint module for in-vehicle voiceprint drift detection. The module may be configured for determining an archived voiceprint for an operator of a vehicle, generating a first real-time voiceprint according to in-vehicle utterances of the operator made while operating the vehicle, generating a first voiceprint deviation to quantify statistical distance between one or more probabilistic density functions associated with each of the first real-time and archived voiceprints, and updating one or more vocal characteristics of the archived voiceprint to generate an updated voiceprint in response to the first voiceprint deviation surpassing an update threshold.