Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training a machine learning model using a deterministic data pipeline. One of the methods may include receiving a first request to generate a deterministic training dataset: transforming raw training examples obtained from the raw data source into pre-processed training examples; assigning a unique index to each pre-processed training example; and caching the pre-processed training examples into the cache directory specified in the received first request; receiving a second request to use the deterministic training dataset to train a machine learning model, the second request specifying a start index; and in response to receiving the second request: reading, from the cache directory, the pre-processed training examples that have indices beginning from the start index; and providing the read training examples in an order of the assigned indices for use in training the machine learning model.

Abstract: A battery housing assembly for housing battery components of an electric vehicle includes a battery tray, a lid, and a sealant. The battery tray includes a channel located outwardly relative to the battery components. The lid is secured to the battery tray and includes an end portion disposed within the channel. The lid further includes at least one alignment feature configured to align the lid and the battery tray. The sealant is disposed within a portion of the channel of the battery tray and is configured to secure the battery tray and the lid to each other. The sealant is further configured to seal an internal cavity of the battery housing assembly.

Abstract: A method of facilitating first mile/ last mile transfer of a vehicle includes: analyzing a first route and a second route using a trained machine learning model to determine, respectively, a first cost of the first route and a second cost of the second route to a user; wherein the trained machine learning model is trained using at least one of a detour in time, a delay in miles, a time of day, a traffic congestion factor, or profile information about a ride share passenger; causing a message including the first cost and the second cost to be displayed to the user via an electronic device of the user; and receiving an indication of whether the user selects the first route or the second route.

Abstract: Described herein are tricyclic PARP1 inhibitors and pharmaceutical compositions comprising said inhibitors.

Abstract: The present invention relates to novel anti-PD1 and anti-LAG3 antibody molecules. The invention also relates to nucleic acids encoding such antibody molecules; to methods for preparing such antibody molecules; to host cells expressing or capable of expressing such antibody molecules; to compositions comprising such antibody molecules; and to uses of such antibody molecules or such compositions, in particular for therapeutic purposes in the field of cancer diseases.

Abstract: An adaptive mapping (AM) computing device having at least one processor in communication with at least one memory device is provided. The AM computing device may be configured to retrieve a plurality of tasks associated with a user, and retrieve geographic mapping data. The AM computing device may also generate a route model based upon the retrieved plurality of tasks and the retrieved geographic mapping data. The AM computing device may execute the route model to determine an optimal route, and transmit, to the user, an optimized travel plan based upon the optimal route.

Abstract: An air conditioner unit includes a cabinet including a base pan that defines at least one alignment feature and a compressor mounted to the base pan using mechanical fasteners that pass through mounting feet and into mounting bosses defined on the base pan. A damper is positioned between the compressor and the base pan and includes a lower pad seated on the base pan, an upper pad extending from the lower pad along the vertical direction and contacting a bottom surface of the compressor, and at least one stopping feature defined on the lower pad or the upper pad, the at least one stopping feature engaging the at least one alignment feature to prevent rotation of the damper.

Abstract: The following relates generally to providing virtual reality (VR) alerts to a driver of an autonomous vehicle. For example, a vehicle may be driving autonomously while the driver is watching a VR movie (e.g., on a pair of VR goggles); the driver may then receive a VR alert recommending that the driver take control of the vehicle (e.g., switch the vehicle from autonomous to manual mode). The following also relates to generating a VR feed for presenting real-time road conditions so that a user may preview a road segment. The following also relates to generating a VR feed corresponding to an event (e.g., a vehicle collision, a crime, a weather event, and/or a natural disaster).

Abstract: A vehicle battery tray includes a floor assembly that has elongated tray sections that attach together at lateral edge portions of adjacent tray sections to form a floor structure that supports vehicle batteries contained in the vehicle battery tray. The elongated tray sections each have a substantially consistent cross-sectional shape that extends longitudinally along a length of the respective tray section. A first tray section includes a first attachment feature at the lateral edge portion thereof and a second tray section includes a second attachment feature at the lateral edge portion thereof that corresponds with the first attachment feature. The first and second attachment features engage together at a longitudinal seam between the lateral edge portions of the first and second tray sections, such as to align upper surfaces of the first and second tray sections.

Abstract: Systems and methods for preventing rotation of an ESP motor when the motor is not powered on, thereby preventing the motor from acting as a generator when fluid flowing through the pump section of the ESP applies a torque to the motor. In one embodiment, an ESP has a motor section, a pump section. The ESP may include a directional coupling that allows unidirectional rotation between the motor shaft and a pump shaft of the pump section, and a directional lock that allows unidirectional rotation between the motor shaft and a housing of the motor section. The directional coupling and directional lock allow the pump shaft to freewheel in the forward direction without causing the motor shaft to rotate, and prevent the pump shaft and motor from rotating in the reverse direction.

Abstract: Techniques of contextual data analysis in distributed computing systems are disclosed herein. In one example, upon receiving operational data and attribute data representing attributes of multiple entities in the distributed computing system, a decision tree is generated based on the received operational and attribute data via machine learning. The decision tree has a root and multiple branches each representing a set of the attributes in the attribute data and a corresponding probability value representing a likelihood that one of the multiple components with the set of the attributes would be associated with an event in the distributed computing system. The multiple branches can then be parsed to identify a common subset of the attributes of the multiple components as most closely related to an occurrence of the event in the distributed computing system.

Abstract: A personalized insurance (“PI”) computing device for determining an optimal insurance product for a driver operating a vehicle for a transportation network company (“TNC”) during a period of increased demand includes at least one processor in communication with at least one memory. The processor is configured to: (i) receive, from a TNC, data indicating increased demand for transportation services, (ii) retrieve driver data that includes the driver history, (iii) generate an optimal pricing model for the driver based upon the increased demand and the driver data, (iv) execute the model to determine an optimal insurance product having characteristics reflecting at least one risk factor associated with the increased demand for transportation services and a risk profile determined from analyzing the driver data, and (v) transmit an offer to the driver to provide transportation services at an increased earnings rate and with the determined optimal insurance product.

Abstract: Devices, systems, methods and computer-readable media are provided for data communication to and from a vehicle. A device is provided that includes memory storing processor-executable instructions; and at least one processor in communication with the memory that executes the stored instructions to: receive, from at least one user on the vehicle, at least one request for data communication; identify a plurality of communication links available at a current location of the moving vehicle; form an adaptive bonded communication link using the plurality of communication links to aggregate throughput across the plurality of communication links for the requested data communication, wherein the adaptive bonded communication link is configured to adapt to data communication requirements for the requested data communication and to data communication characteristics of the plurality of communication links as the vehicle moves.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training a machine learning model using a deterministic data pipeline. One of the methods may include receiving a first request to generate a deterministic training dataset: transforming raw training examples obtained from the raw data source into pre-processed training examples; assigning a unique index to each pre-processed training example; and caching the pre-processed training examples into the cache directory specified in the received first request; receiving a second request to use the deterministic training dataset to train a machine learning model, the second request specifying a start index; and in response to receiving the second request: reading, from the cache directory, the pre-processed training examples that have indices beginning from the start index; and providing the read training examples in an order of the assigned indices for use in training the machine learning model.

Abstract: A pumping system includes a motor and a drive shaft configured for rotation by the motor. The pumping system includes an upper pump positioned above the motor, an upper pump shaft and an upper directional coupling connected between the drive shaft and the upper pump shaft. The upper directional coupling is configured to lock the upper pump shaft to the drive shaft when the drive shaft is rotated in a first direction. The pumping system further includes a lower pump positioned below the motor, a lower pump shaft, and a lower directional coupling connected between the drive shaft and the lower pump shaft. The lower directional coupling is configured to lock the lower pump shaft to the drive shaft when the drive shaft is rotated in a second direction.

Abstract: A handle is coupled to an industrial machine tool, along with a method of operating the industrial machine tool via the handle. The handle comprises a housing, a first switch, and a second switch. The housing includes a first side, a second side, a top end, and a bottom end, the first side facing toward the industrial machine tool and the second side facing away from the industrial machine tool. The first switch actuates a float mode for the industrial machine tool, the float mode being a mode in which movement of a moveable portion of the industrial machine tool is power assisted for an operator of the industrial machine tool. The second switch actuates a machining portion, coupled to the moveable portion of the industrial machine tool, to perform a machining process on a workpiece.

Abstract: A matching computer system for electronically generating, matching, and providing online user profiles, and determining a trust score for a user based upon at least social media data and insurance data is provided. The matching computer system may be configured to register users within the matching computer system, receive consent from the users to capture the social media data, and collect the social media data and the insurance data from each registered user. The matching computer system may also be configured to retrieve the social media data and the insurance data associated with each registered user. The matching computer system may be further configured to determine a trust score for each registered user based upon each respective social media data and each respective insurance data. Each trust score represents a level of trustworthiness of the user.

Abstract: A single-package air conditioner unit may include a housing, an outdoor heat exchanger assembly, an indoor heat exchanger assembly, a compressor, and a plenum. The housing may define an outdoor portion and an indoor portion. The plenum may be attached to the housing and receivable within a wall channel defined by a structure wall. The plenum may define a primary air channel and a make-up air (MUA) inlet in fluid parallel with the primary air channel. The primary air channel may be disposed in fluid communication with the outdoor portion to permit air therebetween. The MUA inlet may be in fluid communication with the indoor portion to permit air thereto.

Abstract: A method of manufacturing a thin film optical apparatus includes providing a substrate and applying an alignment layer over the substrate. The alignment layer ranges from about 50 to 100 nm in thickness. The method includes imprinting a hologram with a desired optic pattern onto the alignment layer and applying at least one layer of mesogen material over the alignment layer.

Abstract: The addition of a selective, fast-acting, short half-life CDK 4/6 inhibitor in a very specific dosage regimen to the combination of chemotherapy with a checkpoint inhibitor provides superior results in the treatment of a tumor or cancer. The unexpected discovery is that the short pulsatile specifically-timed administration of a selective, fast-acting, short half-life CDK 4/6 inhibitor during administration of the chemotherapy portion of the triple combination therapy has a profound effect on the immune cells in the cancer microenvironment.