Abstract: The present disclosure relates to systems and methods for using language models in locations with limited network connectivity. The systems and methods include a hierarchical edge architecture with a plurality of language models with diverse compute capabilities. The systems and methods dynamically select a language model from the plurality of language models to use to respond to a query received by a user in response to determining a level of network connectivity available at a user device.

Abstract: The present disclosure relates to systems and methods for flow scheduling of machine learning workloads. The systems and methods use prior knowledge of the traffic patterns for machine learning workloads and the network topology to schedule the flows of the machine learning workloads. A centralized controller leverages the knowledge of the traffic patterns for the machine learning workloads to guide per-flow routing decisions for the machine learning workflows to schedule the flows across all available network paths to fully utilize network bandwidth capacity.

Abstract: Methods, systems, and computer storage media for providing an indication of an integrity of an object based on a non-invasive assessment of the integrity of the object using acoustic signature management engine in object integrity sensing system. In operation, an aggregate object-intermediate-medium sound of an object in an intermediate medium is detected (e.g., via sensors). An acoustic signature of the aggregate object-intermediate-medium sound is generated as a processed acoustic channel associated with statistical measurements. A reference acoustic signature of the object and intermediate medium is accessed. The reference acoustic signature is associated with an acoustic signature computation model, that generates reference acoustic signatures based on a mean and standard deviation measurements of input signals transmitted through the object and intermediate medium.

Abstract: This disclosure provides a data-driven and scalable method to discover cause-and-effect relationships in data from natural systems that include sparse data sets. This technique can learn a causal graph from heterogenous data sources by combining embeddings from real data and embeddings from simulated data generated by process-based models. The causal graph is used for what-if analysis in out-of-distribution settings. One application is understanding the factors that affect soil carbon. A causal model created by these techniques can be used to discover cause-and-effect relationships that affect soil carbon. This model has applications such as forecasting soil carbon for a future time point to help inform farm practices. Farm practices, like tilling, may be modified in response to predictions provided by the model.

Abstract: A computerized method performs fusion of segmentation masks to generate a refined segmentation mask. A plurality of segmentation masks, each including a group of segments, is obtained for a geographic area. For each segmentation mask, a subgroup of segments is filtered from the group of segments of the segmentation mask based on areas of the subgroup of segments. Pairs of segments of the filtered subgroup of segments are matched to form matched segment groups in the geographic area. Representative segments of the geographic area are selected from the matched segment groups to generate a refined segmentation mask including the selected representative segments. A segment-specific action is performed on at least one of the selected representative segments of the generated refined segmentation mask of the geographic area. In some examples, the segment-specific action is for precision farming in agriculture domain.

Abstract: A computing system for interactive prompting for a supply chain includes processing circuitry that constructs a knowledge graph based ontologies from a plurality of data sources, the ontologies being related to a product. In a turn-based dialog session, the processing circuitry receives a prompt for the product, identifies at least one ontology-level node in a first layer of the knowledge graph, and generates one or more sub-questions. The processing circuitry outputs the sub-questions via a large language model, receives responses to the sub-questions, identifies one or more second-level nodes in a second, middle layer of the knowledge graph based on the responses, and performs a multi-hop query to identify one or more instance-level nodes in the third layer of the knowledge graph. The processing circuitry outputs, via the large language model, text data corresponding to the instance-level nodes as an answer to the prompt.

Abstract: Synthetic molecular tags are placed on an item at various points in a supply chain to create a molecular record of movement through the supply chain. Associations between each unique synthetic molecular tag and individual locations in the supply chain are stored in an electronic record which may be maintained in the cloud. The synthetic molecular tags are collected from the item and sequenced to determine movement of the item through the supply chain by reference to the electronic record. The synthetic molecular tags can be used for identifying recalled items based on locations in the supply chain associated with a recall. The synthetic molecular tags may be polynucleotides such as deoxyribose nucleic acid (DNA). The item may be any type of item including food.

Abstract: Examples are provided that relate to satellites utilizing both a narrowband communication channel and a broadband communication channel. One example provides a satellite comprising narrowband communication hardware configured to communicate on a control plane over a narrowband communication channel using an omnidirectional antenna. The satellite further comprises broadband communication hardware configured to communicate on a data plane over a broadband communication channel using a beamforming antenna.

Abstract: This disclosure introduces a novel method and system for using a large language model (LLM) to create a convenient interface for a complex database. The system includes a custom prompt generator that creates custom prompts from natural language queries. The custom prompts are used to control how the LLM interacts with a database look-up tool. The database look-up tool provides queries to the database in a format understandable by the database and receives responses from the database. This system is useful for obtaining information that is not in a natural language, and thus, is poorly suited for being processed as an embedding by the LLM. Information obtained from the database is included in an answer produced by the LLM.

Abstract: A satellite is provided, including an onboard computing device. The onboard computing device may include a processor configured to receive training data while the satellite is in orbit. The processor may be further configured to perform training at a machine learning model based at least in part on the training data. The processor may be further configured to generate model update data that specifies a modification made to the machine learning model during the training. The processor may be further configured to transmit the model update data from the satellite to an additional computing device.

Abstract: A method for network data communication includes, at a terrestrial computing device, detecting one or more beacon signals from a corresponding one or more orbital communication satellites of a constellation of orbital communication satellites. A transmission probability threshold is adjusted based at least in part on a detected quantity of the one or more orbital communication satellites. Data is transmitted from the terrestrial computing device to the constellation of orbital communication satellites contingent on the transmission probability threshold being satisfied.

Abstract: The present disclosure relates to systems and methods for using language models in locations with limited network connectivity. The systems and methods include a hierarchical edge architecture with a plurality of language models with diverse compute capabilities. The systems and methods dynamically select a language model from the plurality of language models to use to respond to a query received by a user in response to determining a level of network connectivity available at a user device.

Abstract: Synthetic molecular tags are placed on an item at various points in a supply chain to create a molecular record of movement through the supply chain. Associations between each unique synthetic molecular tag and individual locations in the supply chain are stored in an electronic record which may be maintained in the cloud. The synthetic molecular tags are collected from the item and sequenced to determine movement of the item through the supply chain by reference to the electronic record. The synthetic molecular tags can be used for identifying recalled items based on locations in the supply chain associated with a recall. The synthetic molecular tags may be polynucleotides such as deoxyribose nucleic acid (DNA). The item may be any type of item including food.

Abstract: A data processing system implements transmitting an RF signal using a transmitter disposed at a first side of a produce container containing produce to be monitored for quality. The signal is transmitted on multiple frequencies. The system further implements receiving the signal using a receiver disposed at a second side of the produce container opposite the first side of the produce container so the signal passes through the produce; obtaining a sample signal output by the receiver responsive to receiving the signal that passed through the produce contained in the produce container; analyzing the sample signal to identify differences between the RF signal and the sample signal representative of the dielectric properties of the produce; determining an estimated quality level of the produce based on the differences between the RF signal and the sample signal; and outputting an indication of the estimated quality level of the produce.

Abstract: A deep learning system is used to predict crop characteristics from inputs that include crop variety features, environmental features, and field management features. The deep learning system includes domain-specific modules for each category of features. Some of the domain-specific modules are implemented as convolutional neural networks (CNN) while others are implemented as fully-connected neural networks. Interactions between different domains are captured with cross attention between respective embeddings. Embeddings from the multiple domain-specific modules are concatenated to create a deep neural network (DNN). The prediction generated by the DNN is a characteristic of the crop such as yield, height, or disease resistance. The DNN can be used to select a crop variety for planting in a field. For a crop that is planted, the DNN may be used to select a field management technique.

Abstract: The techniques disclosed herein enable systems to optimize generation and dispatch of renewable energies using data-driven models. In many contexts, a renewable energy system is collocated with a local consumer such as a datacenter, a smart building, and so forth. The objective of the renewable energy system is to meet local power needs while participating in various energy markets of differing trading frequencies. To optimally manage the renewable energy system, a data-driven model is configured to analyze current conditions and generate policies to control renewable energy system operations. For instance, the model can retrieve current market prices, generation capacity, costs associated with generating energy, and so forth. Based on the collected information, the model can generate a policy that maximizes revenue obtained by the renewable energy system while meeting local demand. Through many iterations, the model can determine a realistically optimal policy for managing the renewable energy system.

Abstract: The disclosure described herein configures a client device for communication using dynamic spectrum access within a frequency spectrum, such as television white space (TVWS), using a determined location of the client device based on location information, such as from a global positioning system. A dynamic spectrum access database of channels is accessed based on the location information. Available channels are determined for the client device from the channels based on the location information. A list of the available channels for use by the client device are transmitted to the client device, thereby allowing narrowband communication over the channels.

Abstract: Recycling information is associated with objects through the use of molecular tags. The recycling information may describe the type of material that the object is made from as well as provide instructions for recycling. The molecular tags may be polynucleotides or other types of molecules including inorganic molecules. The molecular tags may be embedded within the object or attached to the surface of the object. At the end of the object's life, the molecular tags are read and the recycling information is used to appropriately recycle the object.

Abstract: A data processing system implements receiving telemetry data from a plurality of nodes of a cloud-based computing environment; analyzing the telemetry data using a communication graph pipeline to generate a communication graph representing communication among the plurality of nodes of the cloud-based computing environment; analyzing the communication graph using a role inference pipeline to infer roles of the plurality of nodes of the cloud-based computing environment included in the communication graph and output inferred roles for the plurality of nodes; and performing one or more actions on the communication graph based on the inferred roles for the plurality of nodes.

Abstract: The present disclosure relates to systems and methods for using language models in locations with limited network connectivity. The systems and methods include a hierarchical edge architecture with a plurality of language models with diverse compute capabilities. The systems and methods dynamically select a language model from the plurality of language models to use to respond to a query received by a user in response to determining a level of network connectivity available at a user device.