Abstract: A semiconductor device and fabrication method are described for integrating a nanosheet transistor with a multimodal transistor (MMT) in a single nanosheet process flow by processing a wafer substrate to form buried metal source/drain structures in an MMT region that are laterally spaced apart from one another and positioned below an MMT semiconductor channel layer before forming a transistor stack of alternating Si and SiGe layers in an FET region which are selectively processed to form gate electrode openings so that a first ALD oxide and metal layer are patterned and etched to form gate electrodes in the transistor stack and to form a channel control gate electrode over the MMT semiconductor channel layer, and so that a second oxide and conductive layer are patterned and etched to form a current control gate electrode over the MMT semiconductor channel layer and adjacent to the channel control gate electrode.

Abstract: In general, the disclosure describes techniques for characterizing a dynamical system and a neural ordinary differential equation (NODE)-based controller for the dynamical system. An example analysis system is configured to: obtain a set of parameters of a NODE model used to implement the NODE-based controller, the NODE model trained to control the dynamical system; determine, based on the set of parameters, a system property of a combined system comprising the dynamical system and the NODE-based controller, the system property comprising one or more of an accuracy, safety, reliability, reachability, or controllability of the combined system; and output the system property to modify one or more of the dynamical system or the NODE-based controller to meet a required specification for the combined system.

Abstract: A difference recognition method and a difference recognition system that correctly recognizes the latest progress status. A difference recognition system recognizes the latest progress status of a construction site from a difference, which indicates progress of a construction site. The difference recognition system includes a data management unit that acquires first captured-image data captured in a first state at a first time point and second captured-image data captured in a second state at a second time point that is subsequent to the first time point, and a work determination unit serving as a difference recognition unit that recognizes a newly installed building element from the first captured-image data and second captured-image data as a difference between the first state and the second state. The task determination unit determines the latest progress status from the recognized difference.

Abstract: A magnetoresistive random access memory (MRAM) array includes MRAM cells, each MRAM cell having a corresponding Magnetic Tunnel Junction (MTJ) capable of being in a blown state or non-blown state, in which the blown state corresponds to a permanent breakdown of a tunnel dielectric layer of the corresponding MTJ. Write circuitry performs a one-time-programmable (OTP) write operation to blow selected MRAM cells. For each MRAM cell being blown, the write circuitry uses an initial OTP program reference for the MRAM cell being blown to detect onset of tunnel dielectric breakdown after application of each OTP write pulse of the OTP write operation. After detection of the onset, the write circuitry updates the initial OTP program reference, applies at least one additional OTP write pulse to the MRAM cell being blown, and uses the updated OTP program reference to verify that current saturation of the MRAM cell being blown has occurred.

Abstract: Infusion systems, infusion devices, and related operating methods are provided. In one example, a computer-implemented method includes identifying, based on an input meal size category of a meal event and a plurality of groups of historical meal events associated with a plurality of meal size categories for a person, a group of historical meal events having an associated meal size category that corresponds to the input meal size category, determining at least one person-specific variable representative of historical meal events in the identified group of historical meal events, determining a bolus insulin dosage based at least in part on the at least one person-specific variable, and causing administration of insulin to the person based on the bolus insulin dosage.

Abstract: A method for dynamic estimation of time-to-completion may include: receiving historical ticket data for historical tickets, the historical ticket data comprising historical tasks for each historical ticket; assigning each of the historical tasks into one of a plurality of buckets based on a common feature; training, for each bucket, a trained machine learning engine to predict a time-to-completion for the tasks in the bucket; receiving a current ticket; determining a plurality of tasks and an order of executing the tasks for the current ticket; identifying one of the buckets for each of the tasks in the current ticket; predicting, for each task in the current ticket, the time-to-completion for the current using the trained machine learning engine for the bucket for the task; combining the time-to-completion for the tasks for the current ticket into a current ticket time-to-completion; and returning the current ticket time-to-completion for the current ticket.

Abstract: A computing system is configured to obtain a video that includes text elements and visual elements. The computing system is further configured to generate a plurality of text tokens representative of audio spoken in the video and a plurality of frame tokens representative of one or more frames of the video. The computing system is further configured to generate a set of features that includes a text feature, a frame feature, and a multi-modal feature, wherein the multi-modal feature is representative of multi-modal elements of the video, and wherein generating the set of features is based on the plurality of text tokens and the plurality of frame tokens. The computing system is further configured to associate the set of features with one or more labels to generate a multi-label classification of the video. The computing system is further configured to output an indication of the multi-label classification of the video.

Abstract: Disclosed are techniques related to determination of adjustments to fluid delivery settings. The techniques may involve accessing an input meal size category, where the input meal size category is selected from among a plurality of meal size categories via a user interface; accessing a plurality of groups of historical meal events for a person, where each group of the plurality of groups is associated with a respective meal size category of the plurality of meal size categories and includes respective historical meal events for the person categorized in the respective meal size category; identifying a group, in the plurality of groups of historical meal events for the person, having an associated meal size category that matches the input meal size category; identifying a person-specific carbohydrate amount representative of the historical meal events in the identified group; and determining a bolus dosage value of insulin based on the person-specific carbohydrate amount.

Abstract: Techniques disclosed herein relate generally to diabetes therapy management. In some examples, the techniques involve obtaining therapy-related data (e.g., including sensor glucose data and meal data) of a user of an insulin delivery device for a time period of a plurality of time periods, determining whether the therapy-related data associated with the time period is well-fit data by determining whether a physiological model for the user can be fitted to the therapy-related data associated with the time period, determining one or more parameters of the physiological model by fitting the physiological model to well-fit data associated with the time period, and causing the insulin delivery device to deliver insulin to the user based on the one or more parameters of the physiological model.

Abstract: A system includes one or more processors and one or more processor-readable storage media storing instructions which, when executed by the one or more processors, cause performance of obtaining an insulin-on-board (IOB) value, wherein the IOB value is a quantity of insulin remaining in a patient following delivery of a bolus of insulin, scaling a total daily dose (TDD) value by a factor, to generate a result having a value that is less than the TDD value, based on determining that the IOB value is greater than the result, identifying a condition in which the patient's glucose level will continue to decrease after suspension of basal insulin dosage delivery, and responsive to identifying the condition, causing performance of an action for preventing the patient's glucose level from falling into a hypoglycemic range in combination with the suspension of basal insulin dosage delivery.

Abstract: Techniques disclosed herein relate to configurable target glucose values. In some embodiments, the techniques may involve modifying a glucose setpoint from a first target value to a second target value, wherein the second target value is lower than the first target value. The techniques may further involve regulating a glucose level of a patient to the modified glucose setpoint by controlling insulin delivery. The techniques may further involve after a time period has elapsed, reverting the glucose setpoint to the first target value.

Abstract: This disclosure describes machine learning techniques for capturing human knowledge for performing a task. In one example, a video device obtains video data of a first user performing the task and one or more sensors generate sensor data during performance of the task. An audio device obtains audio data describing performance of the task. A computation engine applies a machine learning system to correlate the video data to the audio data and sensor data to identify portions of the video, sensor, and audio data that depict a same step of a plurality of steps for performing the task. The machine learning system further processes the correlated data to update a domain model defining performance of the task. A training unit applies the domain model to generate training information for performing the task. An output device outputs the training information for use in training a second user to perform the task.

Abstract: A magnetoresistive random access memory (MRAM) array includes MRAM cells, each MRAM cell having a corresponding Magnetic Tunnel Junction (MTJ) capable of being in a blown state or non-blown state, in which the blown state corresponds to a permanent breakdown of a tunnel dielectric layer of the corresponding MTJ. Write circuitry performs a one-time-programmable (OTP) write operation to blow selected MRAM cells. For each MRAM cell being blown, the write circuitry uses an initial OTP program reference for the MRAM cell being blown to detect onset of tunnel dielectric breakdown after application of each OTP write pulse of the OTP write operation. After detection of the onset, the write circuitry updates the initial OTP program reference, applies at least one additional OTP write pulse to the MRAM cell being blown, and uses the updated OTP program reference to verify that current saturation of the MRAM cell being blown has occurred.

Abstract: Disclosed herein are techniques for communication of information. In some embodiments, the techniques involve receiving, at an intermediate device from a remote device, a request to initiate a first communication session. The techniques may further involve responsive to receiving the request, establishing a second communication session between the intermediate device and a medical device. The techniques may further involve responsive to establishing the second communication session, automatically transmitting, by the intermediate device to the remote device, an acknowledgement that establishes the first communication session between the intermediate device and the remote device. The techniques may further involve streaming updated control parameters to the medical device via the second communication session.

Abstract: Disclosed herein are techniques related to configurable target values. The techniques may involve identifying a target glucose value obtained based on user input. Additionally, the techniques may involve determining a fluid delivery command based on a difference between a measured glucose value and the target glucose value. The techniques may further involve causing fluid delivery by a fluid infusion device in accordance with the fluid delivery command to regulate the measured glucose value to the target glucose value.

Abstract: In an example, a method for adapting a machine learning model includes receiving a digital representation of a three-dimensional (3D) object; learning, using a surrogate model, relationships between a plurality of points on a surface of the 3D object; and generating, using the surrogate model, one or more predictions about fluid properties along the surface of the 3D object.

Abstract: A magnetoresistive random access memory (MRAM) array includes a data array and a sensor array. Each MRAM cell includes a Magnetic Tunnel Junction (MTJ). Each MRAM cell of the data array stores a data bit. A first and second column of the sensor array are connected to form a sensor column which includes sensor cells, each formed by a first MRAM cell in the first column together with a second MRAM cell in the second column along a same word line. Only one of a first MTJ of the first MRAM cell or second MTJ of the second MRAM cell is used as an MTJ of the sensor cell, and drain electrodes of select transistors of the first and second MRAM cells are electrically connected. Read circuitry provides read data from the data array and a sensor output indicative of a rupture state of an MTJ of the sensor array.

Abstract: Fluid injection devices and related systems and operating methods are provided. An exemplary injection device includes an actuation arrangement to facilitate delivering fluid from a container, a user interface element coupled to the actuation arrangement, a data storage element to maintain one or more patient-specific parameter values and historical bolus data, and a control module coupled to the actuation arrangement and the data storage element. The control module of the injection device determines a bolus amount of the fluid to be delivered based at least in part on the one or more patient-specific parameter values and the historical bolus data in response to receiving a user input indicative of a desire to administer a bolus, and the control module configures the actuation arrangement to automatically deliver the bolus amount in response to manual actuation of the user interface element.

Abstract: Disclosed herein are techniques related to automatically adjusting at least one parameter of an insulin delivery controller of an insulin delivery device to regulate delivery of insulin to a user. In some embodiments, the techniques may involve obtaining therapy-related data associated with operation of the insulin delivery device for a number of days in the past. The therapy-related data including sensor glucose data for the user and meal data for the user. The techniques may also involve determining at least one adjusted parameter for the insulin delivery controller based on the therapy-related data. Additionally, the techniques may involve causing the insulin delivery controller to adjust at least one setting in accordance with the determined at least one adjusted parameter.

Abstract: Techniques related to temporary setpoint values are disclosed. The techniques may involve causing operation of a fluid delivery device in a closed-loop mode for automatically delivering fluid based on a difference between a first setpoint value and an analyte concentration value during operation of the fluid delivery device in the closed-loop mode. Additionally, the techniques may involve obtaining a second setpoint value. The second setpoint value may be a temporary setpoint value to be used for a period of time to regulate fluid delivery, and the second setpoint value may be greater than the first setpoint value. The techniques may further involve causing operation of the fluid delivery device for automatically reducing fluid delivery for the period of time based on the second setpoint value.