Automated visualization, identification, tracking, and prediction of space objects

Abstract

A system can receive, via a graphical user interface, a request to provide space domain awareness (SDA) information, transmit the request to a machine learning model system to cause a machine learning model to generate a response, wherein the response comprises text information and visualization information, receiving the response from the machine learning model, generating, based on the visualization information, a visualization responsive to the request; and updating the graphical user interface to display the text information and the visualization.

Description

EXOAN.031WO PATENTAUTOMATED VISUALIZATION, IDENTIFICATION, TRACKING, AND PREDICTION OF SPACE OBJECTSINCORPORATION BY REFERENCE OF RELATED APPLICATIONS

[0001] This application claims the benefit under 35 U.S. C. § 119(e) of U.S. Provisional Patent Application No. 63 / 684274, filed August 16, 2024, the entire contents of which are incorporated by reference herein and are made a part of this specification.STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT

[0002] This invention was made with Government support under FA864924P0907 awarded by AFRL SBRK. The Government has certain rights in the invention.BACKGROUNDField

[0003] This disclosure relates generally to tracking space objects such as satellites and automated visual interfaces and computer configurations used in such tracking.Description of Related Art

[0004] Visualization interfaces can be used to allow a user to view, manipulate, and adjust data representing tracked orbital objects (e.g., satellites). Tracking orbital objects involves taking in an amount of data and incorporating that data into a workable and usable interface. Tracking orbital objects may be done using photographs of objects in space and tracking their positions using a plurality of photographs. Visualization systems have been developed in various fields that provide some functionality with regard to portraying various information. However, many features are lacking for effective interaction with a user and many problems exist in various aspects of user access to space domain awareness data for which this application provides solutions.SUMMARY

[0005] Example embodiments described herein have innovative features, no single one of which is indispensable or solely responsible for their desirable attributes. Without limiting the scope of the claims, some of the advantageous features will now be summarized.

[0006] In some aspects, the techniques described herein relate to a system including: a computer-readable memory; and one or more processors programmed by executable instructions to at least: receive, via a graphical user interface, a request to provide space domain awareness (SDA) information; transmit the request to a machine learning model system to cause a machine learning model to generate a response, wherein the response includes text information and visualization information; receiving the response from the machine learning model; generating, based on the visualization information, a visualization responsive to the request; and updating the graphical user interface to display the text information and the visualization.

[0007] In some aspects, the techniques described herein relate to a computer-implemented method including: receiving, from a user interface system, a request to provide space domain awareness (SDA) information; generating a vector representation of the request; generating a result of a vector search of an SDA database based on the vector representation, wherein the SDA databased includes a plurality of SDA information items, wherein the SDA information items of the SDA database are stored in a vector form, wherein the vector search is based on determining a vector similarity between the vector representation of the request and the vector format of an SDA information item, and wherein the result of the vector search is at least one SDA information item in vector form having a closest vector similarity to the vector representation; applying the result and the vector representation as input to a machine learning model to cause the machine learning model to generate a natural language response to the request, wherein the response includes at least a portion of the result, wherein the result further includes visualization information associated with the at least one SDA information item of the result; and causing display of the response based on transmitting the response to the user interface system, wherein causing display of the response includes generating a visualization associated with the at least one SDA information item of the result based on the visualization information.BRIEF DESCRIPTION OF THE DRAWINGS

[0008] The following drawings and the associated descriptions are provided to illustrate embodiments of the present disclosure and do not limit the scope of the claims.

[0009] Figure 1A schematically illustrates an example system for extracting space domain awareness data (SDA) data from raw space domain data and presenting the SDA data to a user in response to requests received via a graphical user interface (GUI) or a text-based artificial intelligence (Al) assistant in natural human language.

[0010] Figure IB schematically illustrates an example visualization system that may be used in the system shown in Figure 1 for receiving and processing raw data and visualizing the processed data.

[0011] Figure 2 schematically illustrates an example visualization display that may be generated by the system shown in Figure 1 in response to a request for SDA data received from user via the text-based Al assistant or the GUI of the system shown in Figure 1.

[0012] Figure 3 schematically illustrates an example report that may be generated by the system shown in Figure 1 in response to a request for SDA data received from a user via the text-based Al assistant or the GUI of the system shown in Figure 1.

[0013] Figures 4A-4E illustrate example Al enhanced interfaces for receiving and responding to queries for SDA data received from a user, and displaying text and visualizations responsive to the queries.

[0014] Figures 4F-4G illustrate example user interactions with the text-based Al assistant shown in Figure 4A showing queries for SDA data and response generated, in natural human language.

[0015] Figure 4H illustrates an example chat box (left) and display (right panel) generated by an Al enhanced interface showing queries for SDA data and the requested SDA data presented in the chat box and the display.

[0016] Figure 5 is a block diagram of an illustrative computing system configured to provide automated visualization, identification, tracking, and prediction of space objects according to some embodiments.

[0017] Figure 6 illustrates example data flows within an example environment for providing automated visualization, identification, tracking, and prediction of space objects according to some embodiments.

[0018] Figure 7 illustrates an example interaction between a user requesting information from a system for providing automated visualization, identification, tracking, and prediction of space objects according to some embodiments.

[0019] Figure 8 illustrates example interactions between computing systems for retrieving SDA data as part of providing automated visualization, identification, tracking, and prediction of space objects according to some embodiments.

[0020] Figure 9 schematically illustrates communications between computing systems to enable a chat interface for a system providing automated visualization, identification, tracking, and prediction of space objects according to some embodiments.

[0021] These and other features will now be described with reference to the drawings summarized above. The drawings and the associated descriptions are provided to illustrate embodiments and not to limit the scope of any claim. Throughout the drawings, reference numbers may be reused to indicate correspondence between referenced elements. In addition, where applicable, the first one or two digits of a reference numeral for an element can frequently indicate the figure number in which the element first appears.DETAILED DESCRIPTION

[0022] Although certain embodiments and examples are disclosed below, inventive subject matter extends beyond the specifically disclosed embodiments to other alternative embodiments and / or uses and to modifications and equivalents thereof. Thus, the scope of the claims appended hereto is not limited by any of the particular embodiments described below. For example, in any method or process disclosed herein, the acts or operations of the method or process may be performed in any suitable sequence and are not necessarily limited to any particular disclosed sequence. Various operations may be described as multiple discrete operations in turn, in a manner that may be helpful in understanding certain embodiments; however, the order of description should not be construed to imply that these operations are order dependent. Additionally, the structures, systems, and / or devices described herein may be embodied asintegrated components or as separate components. For purposes of comparing various embodiments, certain aspects and advantages of these embodiments are described. Not necessarily all such aspects or advantages are achieved by any particular embodiment. Thus, for example, various embodiments may be carried out in a manner that achieves or optimizes one advantage or group of advantages as taught herein without necessarily achieving other aspects or advantages as may also be taught or suggested herein.

[0023] Unless explicitly indicated otherwise, terms as used herein will be understood to imply their customary and ordinary meaning.Introduction

[0024] Extracting and presenting space domain awareness (SDA) data needed by a user from raw space domain data can be a challenging task and, in some cases, may require extensive training and familiarity of the user with specialized software tools. For example, an un-trained user (e.g., a non-expert user) may not be able to quickly and efficiently obtain desired information about one or more orbital objects or events associated with the one or more orbital objects from raw space domain data using existing software tools. An objective of the disclosed methods and systems is to facilitate and improve interaction of an un-trained user to efficiently interact with an artificial intelligence (Al) enhance interface so that without using highly specialized tools and procedures the un-trained user can readily obtain desired SDA data that may be required to make a decision or take an action to avoid operational surprises. In some examples, raw space domain data may comprise data (e.g., photographs of satellites) received from one or more satellites, telescopes, or other systems configured to observe or track orbital objects (e.g., satellites). In some examples, the raw space domain data may be received from at least two telescope at two different locations in on earth. In some examples, raw space domain data may include data may comprise names and characteristics of orbital objects. In various implementations, raw space domain data may be received from a satellite or telescope in real time, from a data store that stores historical space domain data received from satellites or telescopes, or from a data base (e.g., a government data based) storing space domain information (e.g., information associated with various satellites). In some embodiments, SDA data may comprise previously processed data generated by processing raw space domain data (e.g., by an SDA analysis system with or without expert assistance). For example, SDA data may comprise telescope or satellites images labeled or augmented withidentifiers (e.g., longitude, latitude, time, name, a scalar identifier, or the like). Further SDA data may comprise estimations, calculations or otherwise results obtained by processing one or more telescope or satellites images. For example, SDA data may include alerts associated with orbital objects and orbital events involving one or more orbital objects, orbits, or data associated with a conjunction of orbital objects, maneuvers performed by orbital objects, estimated parameter values to perform a maneuver, and the like. In various scenarios, SDA data may be generated by an SDA data analysis system, automatically or in response to inputs or queries received from a user (e.g., from a trained user and / or an expert user). In some embodiments, SDA data may comprise labeled data generated by an expert user using an SDA data analysis system. In some examples, the SDA analysis system can be a system (e.g., a software suite ruing on a local or cloud placed computing platform) configured to receive raw space data and generate SDA data based on a user interaction with a specialized GUI (e.g., including tools that may require training and background knowledge to be used effectively). In some embodiments, SDA data automatically generated by the SDA data analysis system or an expert using the SDA data analysis system may be stored in a space domain Unified Data Library (UDL) for future use by the SDA data analysis system or another system. As such, the UDL may comprise large SDA data sets generated using raw space domain data obtained during long periods of orbital observation. In some embodiments, SDA data stored in a space domain UDL may be useful for a variety of applications including but not limited to tracking and controlling space objects, without further processing. Some of the disclosed methods and system may allow a user (e.g., a non-expert user) to obtain desired SDA data stored in a space domain UDL.

[0025] In some cases, large and complex raw space domain datasets can pose a challenge for existing space domain analytical tools and corresponding methods of analysis and significant time and expertise may be required to extract meaningful insights from raw space domain data (herein referred to as raw data). Current methods and tools (e.g., software tools) used for collecting, labeling and analyzing raw data can be complex, cumbersome, and inefficient and thereby, in some cases, result in potential risks due to a delayed and / or inaccurate action. Some of the disclosed methods and systems may allow a user (e.g., an un-trained and / or non-expert) to avoid such risks by obtaining accurate insights from raw data in a timely fashion.

[0026] Some embodiments disclosed herein may allow an SDA data analysis system or another system (e.g., a cloud-based SDA data search system) to understand and respond to userqueries in natural language and thereby facilitate rapid comprehension and interpretation of raw data. For example, a system may use a large language model (LLM) to respond to user queries in natural language (e.g., received via a chat box). In various implementations, the response may comprise textual information, graphs, or other forms of data presentation which may be selected, adjusted, and / or modified by the user, in some cases, using the same user interface (e.g., chat box) used to receive the user queries. Compared to manual analysis and specialized tools, the proposed systems and tools may empower users to efficiently extract actionable insights from SDA data by providing conversational interaction, scalability, and accessibility.

[0027] In some embodiments, the disclosed systems may synergistically combine some of the existing analytical tools developed for receiving, labeling, and processing raw data, with tools and methods developed for providing text-based user interfaces that comprehend natural human language and interact with a user through a two-way conversation augmented by customized and dynamic visualization displays developed for presenting SDA data. In some embodiments, at least a portion of the system may be implemented in a cloud computing platform to facilitate access, and further to benefit from additional tools (e.g., analytical, data management, LLMs, chatbots, and the like) provided by (or available though) the cloud computing platform. In some cases, the cloud computing platform may ensure robust infrastructure support and allow the system to be easily improved using new tools that may become available on a selected cloud computing platform. In some implementations, the system may use a generative pre-trained transformer (GPT) available through the cloud computing platform to provide a text-based artificial intelligence (Al) assistant for responding to queries in natural language. For example, the system may use Microsoft Azure as the cloud computing platform and ChatGPT as the text-based Al assistant. In some embodiments, the text-based Al assistant may be configured and / or finetuned to access and use SDA datasets previously generated by an SDA analysis system and respond to natural language queries directed to orbital objects and corresponding events and interaction by extracting requested information from the SDA. In some cases, the processed datasets may comprise SDA data including but limited to expertly-labeled data sets obtained from telescope observations, satellite states, critical events such as conjunctions and close approaches substantially in real-time, and the like. In some embodiment, SDA data may be stored in a local space domain Unified Data Library (UDL). In some cases, the space domain UDL (herein referredto as UDL) may be at least partially stored in a cloud computing platform (e g., the same platform that support the text-based Al assistant).

[0028] In some embodiments, the text-based Al assistant may be further configured and / or finetuned to access the analytic tools of an SDA data analysis system and use these tools to generate information requested via the text-based interface using natural language human language. Additionally, the text-based Al assistant may be further configured and / or finetuned to access the presentation and display tools available in the SDA data analysis system and use these tools to present the information requested via the text-based interface based on presentation instructions provided in natural language human language. In some embodiments, an Al system may be configured to respond to queries received via a text-based Al assistant by directly accessing raw data, processing the raw data, generating SDA data, and presenting the SDA data. In some embodiments, the Al assistant may be multimodal, and configured to accept input or generate output in multiple modalities. A modality of the Al assistant may refer to, for example, text, image, video, audio, or multimodal information. Additionally, the Al assistant may be configured to generate instructions for a visualization system to generate a visualization related to a response provided by the Al assistant. In some embodiments, such as where the Al assistant is a multimodal Al assistant, the Al assistant may generate a portion of the visualization. For example, the Al assistant may access or receive orbital path information for an orbital object from an SDA data analysis system and generate an image of the 2-dimensional orbital path of the object. The Al assistant may then provide instructions to a visualization system to generate a 2-dimensional representation of the Earth, and overly the orbital path image on the Earth image to generate a visualization to be provided to a user.Space domain data access and analysis powered by artificial intelligence

[0029] Described herein are methodologies and related systems for processing raw space domain data, herein referred to as raw data, associated with one or more orbital objects (e.g., satellites) and generating displays, reports, alerts, to provide a user with information about the one or more orbital objects and their interactions. In various embodiments, the raw data may be received, among other sources, from telescopes, satellites, or databases, e.g., via a network connection. In some embodiments, raw data may comprise photographs received from one or more satellites. In some embodiments, the system may allow the user to request information aboutorbital objects and modify information presented, and adjust various parameters used by the system to extract information from raw data and present the information. However, the embodiments are not so limited and the system may have other capabilities and provide other functionalities with respect to receiving and selecting raw data, processing raw data, and presenting the resulting information. For example, in some embodiments, the disclosed methods and systems may process the raw data to determine an action or parameters associated with an action associated with controlling an orbital object (e.g., performing a maneuver to reposition or redirect an orbital object). It will be understood that although the description herein is in the context of satellites, one or more features of the present disclosure can also be implemented in tracking objects other than satellites like, for example, aircraft, watercraft, projectiles, and other objects. Some embodiments of the methodologies and related systems disclosed herein can be used with various tracking systems, including, for example, those based on government databases.

[0030] In some embodiments, the disclosed systems may comprise an electronic processor and core machine-readable instructions (e.g., a software suite) executed by the electronic processor. In some embodiments, at least a portion of the core machine-readable instructions may be executed by an electronic processor in cloud computing platform or system. In some such embodiments, the system may use one or more tools available via the cloud computing platform to enhance or augment the performance of the core machine-readable instructions with respect to data processing, data management, user interaction, or other aspects. In some cases, the system and tools implemented based on the core machine-readable instructions may be referred to as the SDA data analysis system or the main system.

[0031] In some embodiments, execution of the core machine-readable instructions by the process or may cause the disclosed systems to generate a graphical user interface (GUI) configured to receive user inputs and present requested information via a visualization display comprising one or more graphs, charts, tables, or other interactive interfaces and displays. In some examples, a trained user may interact with such GUI to select a portion of raw data and cause the system to present, process, and / or modify the received portion of raw data and generate requested information or SDA data. For example, the user may the GUI to generate a graph indicating temporal change of longitude of orbital objects based on raw data received from one or more telescopes during a specified time window and cause the system to identify an orbital object in the graph by selecting the orbital object. As such, in some embodiments, to generate the desiredinformation using the system, a user may need to be familiar with various functions and features of the GUI and additionally with have a minimum level of knowledge about dynamic and kinematic of orbital objects and related features and characteristics.

[0032] Some embodiments disclosed herein may provide an enhanced user interface (e.g., a Al enhanced user interface) to facilitate user interaction and allow a broader range of users having different levels of training and knowledge or having basic knowledge and no-training to effectively interact with the system to obtain information about orbital objects. In some embodiments, the enhanced user interface may be generated using artificial intelligence (Al) and Al models trained using machine learning methods (e.g., deep learning). For example, a trained model may be used as a bridge or translator between the main system, which is controlled and enabled by the core machine-readable instructions, and a non-expert user who may not be trained to use the GUI. The terms machine learning model and Al model may be used throughout the description to represent a model.

[0033] In some embodiments, the system may comprise an Al enhanced interface configured to receive instructions and requests about orbital objects and events and in human language (e.g., provided via a text prompt), and provide the requested information in textual and / or graphical form. In some implementations, the Al enhanced interface may comprise an Al chat bot that receives natural language instructions via a text prompt. In some examples, the Al chat bot may be implemented by fine tuning or adapting a pre-trained large language model (LLM) based on the features of the main system and data generated by the main system. In some implementations, adapting and / or fine tuning the pre-trained LLM may comprise adjusting certain weights and coefficient of the pre-trained LLM to optimize it for understanding natural language requests associated with orbital objects and generating responses comprising orbital objects and related attributes (e.g., satellite name, longitude, latitude, orbits, inclination, speed, and the like), actions (e.g., maneuver), events (e.g., conjunction) or other aspect.

[0034] Figure 1A schematically illustrates an example system 100 for extracting space domain awareness data (SDA) data from raw space domain data and presenting the SDA data to a user in response to requests received via a graphical user interface 105 (GUI), or an interactive chat interface 107 that may include a text-based artificial intelligence (Al) assistant, in natural human language. In some embodiments, the system 100 may be configured to execute machine readable instructions to generate the GUI 105 or the interactive chat interface 107. In someembodiments, the GUI may comprise a visualization display (visualization display-1) comprising one or more sections for presenting graphs, charts, text, and images. In some embodiments, the Al enhanced interface may comprise a chat box (text-based Al assistant) and, in some cases, a visualization display (visualization display-2) configured to display information or data requested via the chat box. It should be understood that elements of the various example systems described in relation to the figures herein are interchangeable and may be combined to provide additional funcitonality, such that functionality described with respect to particular elements in a first figure may be combined with functionality described with respect to a second figure. For example, the Al engine 111 described with respect to FIG. 1 may provide at least a portion of the functionality of the machine learning-based response generation and data access system 610 or the Al engine 614 of FIG. 6.

[0035] In some embodiments, the system may include a main system 109 configured to receive raw data (e.g., photographs) from telescopes, satellites, data bases, or other data sources, via a data network and process the raw data automatically or in response to user interaction with a user interface (e.g., the graphical user interface 105 or the interactive chat interface 107). Raw data may include real-time or substantially real-time telescope data, historical telescope data, or other data (e.g., data received from a government database). In some embodiments, the user interface may comprise a graphical user interface (GUI) 105 or an interactive chat interface 107, e.g., a textbased Al assistant that interacts with the user in natural human language.

[0036] In some embodiments raw data may include image data (e.g., photographs) of portions of the sky from one or more telescopes positioned at various positions across the globe. The photograph data can be used to map out the entirety or near entirety of the sky. Various altitudes above sea level may be tracked. The data can be tracked and processed substantially in real-time. For example, a contemporary database may be configured to receive real-time image data. A historical database may be configured to store data received before a threshold time. The threshold time may be a specified amount of time (e.g., years, months, days, etc.). Alternatively, the threshold time may refer to a time based on a user action. For example, the historical database may be configured to store data received before a user causes the system to display the user interface. Using an algorithm, the received data may be consolidated and categorized. For example, the algorithm may be configured to determine whether objects that appear in a plurality of photographs correspond to the same object over time and space.

[0037] In some embodiments, by processing the raw data the main system 109 may generate processed data, also referred to SDA data, that may include data labeled with identifiers and organized based on the identifiers (e.g., as tracks associated with different orbital objects that may indicate orbital movements of the respective orbital objects), or results of comparisons, calculations or estimation based on raw data or the labeled data (e.g., alerts associated with orbital objects and orbital events involving one or more orbital objects, orbits, or data associated with a conjunction of orbital objects, maneuvers performed by orbital objects, estimated parameter values to perform a maneuver, and the like).

[0038] In some examples, the processed data may be generated at least partially based on inputs and instructions received from a user (e.g., an expert user) via the GUI 105 and in some cases, the interactive chat interface 107. For example, the user may select and label (or relabel) data via the GUI, trigger an orbit estimation, and request alert generation for certain orbital events. In some cases, the main system 109 may present the processed data via the GUI, or the Al enhanced interface. In some cases, the GUI 105 and the interactive chat interface 107 may be presented simultaneously to the user in the same interface, for example as shown in FIGS. 4A-4D.

[0039] In some embodiments, the main system 109 may store at least a portion of the processed data in a data store (e.g., a local data store or cloud-based data store). In some cases, the data store, also referred to as the space domain UDL or UDL, may store a large number of SDA data sets generated over an extended period.

[0040] In some embodiments, the system may comprise an Al platform or Al engine 111 configured to execute machine readable instructions based on a large language model to enable user interaction with the system via the Al enhance interface and using natural human language. In some embodiments, the Al engine 111 receives information from the Al enhanced interface via the main system 109. The main system 109 may perform some initial processing on the user interaction information received from the Al enhanced interface and provide the processed interaction information to the Al engine 111. Alternatively, the main system 109 may provide the user interaction data directly to the Al engine 111. Further, the main system 109 may process the response generated by the Al engine 111. For example, the main system 109 may process visualization instructions from the Al engine 111 to generate a visualization for the Al enhanced interface. In some embodiments, the machine-readable instructions executed by the Al engine 111 may comprise a generative pre-trained transformer (GPT) available. In some embodiments, theGPT may have been configured and / or finetuned to access and use SDA datasets stored in the processed data store or to interact with the main system 109. In some embodiments, the Al platform or Al engine 111 may compromise an interface control configured to generate the Al enhanced interface.

[0041] In one embodiment, in response to receiving a query via the chat box, the Al engine 111 may access the processed data store, search and select data from the processed data store based on the query and present the data via the chat box or the visualization display (visualization display-2). In some embodiments, a portion of the processed data may be transferred to the Al platform and the Al engine 111 may search for requested data in the transferred portion. In some embodiments, the system may comprise an auto-ingestion system to retrieve SDA data from the UDL and transfer it to a database of the cloud computing system.

[0042] In some implementations, in response to a first text prompt provided in the chat box the user may request information and in response to a second text prompt provide a preferred method of presentation (e.g., test, plot, etc.) and / or parameter that may be used to modify a default presentation mode.

[0043] In one embodiment, in response to receiving a query via the chat box, the Al engine 111 may access the processor to the main system 109 to access raw data and request a service or function to process the raw data to generate data requested by a user. In these embodiments, one or more functions and tools that are provided by the main system 109 via the GUI may become available through the Al enhanced interface and based on natural language queries.

[0044] The tools, services, and process data generate by the main system 109 and the Al engine 111 may generally relate to the tracking of objects in orbit (e.g., satellites), other space objects, and providing an interactive user interface to interact with data related to the tracking of these objects.

[0045] In some embodiments, the Visualization display - 1 of the GUI may be displayed on any type of digital display device, such as a desktop computer, a laptop computer, a projection-style device, a smartphone, a tablet, a wearable device, or any other display device. In some cases, the visualization display-1 may include one or more of a first plot longitude-time plot (or graph), a longitude-latitude plot, a scalar plot, an image chip. However, the embodiments are not so limited and in various embodiments, other forms of data or image presentation may used.In some examples, two or more plots in the visualization display-1 may be synchronized. In some embodiments the visualization display-2 may comprise one or more features of the visualization display- 1.

[0046] As described above, in some embodiments, the Al engine 111 or platform may comprise a cloud-based platform that includes an tools for supporting an Al enhanced interface having a chat box configured to receive queries and provide response in natural human language. In some embodiments, the cloud-based platform may comprise pre-trained Large Language Model (LLM) that has been further fine-tuned to be used for SDA data access, analysis, and presentation. Advantageously, integrating an LLM, e.g., using commercially available software packages and systems, such as Microsoft's Azure OpenAI studio, may facilitate intuitive and efficient interaction between users or operators and the vast trove of SDA data.

[0047] In some embodiments, the SDA data stored in the processed data store (e.g., a space domain unified data library, UDL) can be standardized and monitored in centralized manner. For example, the system may persistently monitor space operations in GEO and update the UDL. In some examples, UDL may include a large amount of correlated observations, orbital states, orbital events that verified, delivered, and / or updated substantially in real-time. In some examples, UDL may include real-world Exo data, including observations, curated alerts, satellite launch information, ownership, and mission details.

[0048] In some embodiments, data may be tagged and flagged (e.g., by a raw data analysis and receiving system, e.g., Espoc). In some cases, the system may be configured to enable the streaming of labeled data as a service. In various implementations, the users can benefit from access to SDA data immediately or with a short delay upon data population in the UDL. In some embodiments, the system may support scaled and secure model development using SDA data from UDL for rapid data-driven fine-tuning, configuring, or training, digital models (e.g., pre-trained LLMs) to be used for receiving natural language queries directed to orbital objects, orbital events and generally space domain awareness, and generating response by finding, analyzing, and presenting SDA data.

[0049] In some embodiments, flexible sensor tasking and expert data labeling (e.g., guardian-supervised data labeling) may be used to generate the processed data to enable rapid and responsive iteration for Al tool development.

[0050] In some embodiments, the Al engine 111 may be configured to access the data network, directly or via the main system 109, receive raw data, and process raw data to respond to a query. In some such embodiments, the Al model (e.g., the pre-trained LLM) may be finetuned, adjusted, or further trained to analyze raw data, label raw data, or otherwise use raw data to calculate (or estimate) to respond to requests. As such, in some embodiments, the Al engine 111 may be configured to perform one more functions and services provided by the main system 109 at least partially independent of the main system 109. For example, in response to a query directed to an orbital state of an orbital object the Al engine 111 may receive raw data via the data network, label the raw data to identify data associated with the orbital object, and the present the requested state. In some cases, the Al engine 111 may use the processed data available via the data store to process the raw data received from the data network.

[0051] As such, in various implementations, the Al engine 111 may respond to queries received via the chat box, using one or both of processed data (from the processed data store) and raw data (from the data network), with or without using the main system 109.

[0052] In various implementations, processed data may comprise predicted data, orbital path data metadata stamp, image data, image data derived from photographs, object identifier stamp, latitude stamp, scalar stamp, image chip. In some examples, the scalar may comprise a magnitude, a projected area, a temperature, a mass, a radar cross section, an altitude, an inclination, a delta-v, a time until a certain event, or a probability of a certain event.

[0053] Various embodiments of the system 100 shown in FIG. 1 may enable users to interact with SDA data via text prompt and by integrating advanced machine learning capabilities with modules of the main system 109 and the processed data and outputs generated by the main system 109. Advantageously, by enhancing accessibility, interpretability, and usability, the disclosed embodiments can enhance access to SDA across the space industry such that informed decisions can be made and risks can be mitigated risks more effectively in the increasingly complex orbital environment. Further, allowing user interaction with data in natural language may enable more efficient resolution of user queries by minimizing the number of attempts by a user to cause the system 100 to respond to the user intention represented in the user query.

[0054] Advantageously, integrating continuously updated data into the chat interface may enable users to access the latest information rapidly and thereby avoid operational surprise. Usage of an LLM may enable understanding and responding to user queries in natural language,facilitate rapid comprehension and interpretation of intricate SDA data, and allow users to extract actionable insights from SDA data more efficiently with less computer processing than in previous systems.

[0055] Figure IB schematically illustrates an example visualization system 190 that may be used by the main system 109 shown in Figure 1 for receiving and processing raw data and visualizing the processed data. As illustrated, the visualization system 190 can include a hardware processor 188, a memory 146, a real-time orbital object data interface 172, a tagging interface 174, a image interface 176, a real-time connection interface 178, and / or an real-time connection interface 178, each of which can communicate with one another by way of a communication bus 142 or any other data communication technique. The hardware processor 188 can read and write to the memory 146 and can provide output information for the visualization display-1 of the GUI. The real-time orbital object data interface 172, tagging interface 174, image interface 176, and / or real-time connection interface 178 can be configured to accept input from an input device 164, such as a keyboard, mouse, digital pen, microphone, touch screen, gesture recognition system, voice recognition system, and / or another input device capable of receiving user input. In some embodiments, the visualization display-1 and the input device 164 can have the same form factor and share some resources, such as in a touch screen-enabled display.

[0056] In some embodiments, the real-time orbital object data interface 172, the tagging interface 174, the image interface 176, and / or the real-time connection interface 178 can be connected to a historical data server 140, a contemporary data server 150, and / or a metadata server 154 via one or more networks 144 (such as the Internet, 3G / Wi-Fi / LTE / 5G networks, satellite networks, etc.). The real-time orbital object data interface 172 can receive graphical data information related to orbital objects via the network 144 (the network 144 can provide one-way communication or two-way communication.

[0057] In various embodiments, the Al engine 111 of the system shown in FIG. 1A may have access to one or more interfaces and / or modules of the visualization system 190 to respond to a query received from the chat box. For example, Al engine 111 of the system may use one or more interface modules of the visualization system 190 to receive raw data from the network, and in some cases, label the received raw data and / or use one or modules of the visualization system 190 to process the raw data or the labeled data to generate information requested by a user.

[0058] Figure 2 schematically illustrates visualization display 200 as an example of the visualization display-1 that may be generated by the main system 109 shown in Figure 1 in response to a user request for SDA data received via the text-based artificial intelligence (Al) assistant or the GUI of the system shown in Figure 1. In the example shown, the visualization display 200 comprises a longitude-time graph 204, scalar-time graph 208, a longitude-latitude graph 212, and a display area 216.

[0059] The visualization display 200 can include a longitude-time graph area 228. In some embodiments, the longitude-time graph area 228 is bounded by a first longitude axis 224 and a first time axis 220. Each of the first longitude axis 224 and / or first time axis 220 can include one or more axis labels. The first longitude axis 224 may span any portion of longitudes found on a planet (e.g., Earth). For Earth, the range may be from 180 W (e.g., 180° West or -180°) to 180 E (e g., 180° East or +180°) or any range therein. However, other ranges are possible, examples of which are described below. The first longitude axis 224 may run eastern-most to western-most from left to right, but other configurations are possible.

[0060] The first time axis 220 may span any time from a historical time to nearly a current time of a user. The displayed time may correspond to a universal time, such as the coordinated universal time (UTC). Stored time values may similarly be in UTC. The most current time available may include time within a few seconds (e.g., 1-60 seconds) or a few minutes (e.g., 1-30 minutes) of a present time at which a viewer is observing the data. The first time axis 220 may span from a historical time from an earliest time when a database (e.g., a historical data server 140, a miscellaneous data server 154) has available data. In some embodiments, the historical data server 140, the contemporary data server 150, and / or the metadata server 154, may be configured to store some or all of the corresponding data in short-term memory storage (e.g., Random Access Memory (RAM)). The first time axis 220 may include axis labels that run earliest to most recent from top to bottom, but other variations are possible. An axis label may show a corresponding time to include a year, a month, a day, an hour, a minute, and / or a second, depending on the level of specificity that is available, the span of the first time axis 220, and / or the level of detail that is needed for a particular display.

[0061] To further aid a user in visualizing the orbital object information, in some embodiments, the longitude-time graph 204 may display a longitude-time map. The longitudetime map may be a geographical map of a portion of the planet. For example, the longitude-timemap may identify the contours and / or limits of various landmasses (e.g., continents, islands). This information may help a user quickly ascertain over which landmass or body of water, for example, an orbital object may be located. For example, it may be useful to a viewer to see that a satellite orbits above a portion of Africa (or other planetary location). Points displayed on the corresponding graph (e.g., the longitude-time graph 204) may be superimposed over the geographic map (e.g., the longitude-time map).

[0062] The longitude-latitude graph 212 may include a longitude-latitude graph area 240 that is bounded by a second longitude axis 236 and a latitude axis 232. Each of the second longitude axis 236 and / or the latitude axis 232 can include one or more axis labels. The second longitude axis 236 and the first longitude axis 224 may be identical. In some embodiments, the axis labels of the second longitude axis 236 represent the values of the axis labels for the longitudetime graph 204. Like the first longitude axis 224, the second longitude axis 236 may span any portion of longitudes found on the planet. Like the first longitude axis 224, the second longitude axis 236 may run eastern-most to western-most from left to right, but other configurations are possible.

[0063] The latitude axis 232 may span any latitude found on the planet. For example, the latitude axis 232 may span from 90 S (e.g., 90° South) to 90 N (e.g., 90° North) or any range therein. The latitude axis 232 may include axis labels that run western-most to eastern-most from left to right, but other variations are possible.

[0064] In some embodiments, the longitude-latitude graph 212 may display a longitude-latitude map. In some designs, the longitude-latitude map may include a portion of the same features in the longitude-time map. The longitude-latitude map may be a geographical map of a portion of the planet. For example, the longitude-latitude map may identify the contours and / or limits of various landmasses (e g., continents, islands). This information may help a user quickly ascertain over which landmass or body of water, for example, an orbital object may be located. For example, it may be useful to a viewer to see that a satellite orbits above a portion of Africa (or other planetary location). Points displayed on the corresponding graph (e.g., the longitude-latitude graph 212) may be superimposed over the geographic map (e.g., the longitude-latitude map).

[0065] The scalar-time graph 208 may include a scalar-time graph area 252 that is bounded by a scalar axis 248 and a second time axis 244. Each of the scalar axis 248 and / or the second time axis 244 can include one or more axis labels. The second time axis 244 and the firsttime axis 220 may be identical. For example, first time axis 220 may respond to a user input in the same way as the second time axis 244. In some embodiments, the axis labels of the first time axis 220 represent the values of the axis labels for the scalar-time graph 208. Like the first time axis 220, the second time axis 244 may span any time from a historical time to nearly a current time of a user.

[0066] The scalar axis 248 may span any value of scalars associated with scalars within a database. Each scalar displayed may correspond to a magnitude or other value. For example, the magnitude may represent an intensity (e.g., of light from the orbital object). However, other scalar values are also possible, such as a size, a projected area, a temperature, a mass, a radar cross section, an altitude, an inclination, a delta- V, a time until a certain event, a probability of a certain event, etc.

[0067] The visualization display 200 may further include a display area 216. The display area 216 may be configured to display an image chip 268. This may offer a viewer an opportunity to see an underlying photograph from which image data were extracted that correspond to a set of data or identifiers that are associated with one or more points displayed by the visualization display 200. The image chip 268 may correspond to a photograph of one or more orbital objects. For example, the image chip 268 may be a representation of the photograph. In some cases, the image chip 268 may display an object image 270 that represents an orbital object. The image chip 268 may include multiple object images 270 (e.g., sequential images, summated images (see below), etc.).

[0068] The one or more points displayed by the visualization display 300 may be received from one or more databases (e.g., the historical data server 140, the contemporary data server 150, the metadata server 154) via one or more data interfaces (e.g., the real-time orbital object data interface 172, the tagging interface 174, the image interface 176, the real-time connection interface 178). The data interfaces may be referred to as application program interfaces (e.g., APIs).

[0069] In some embodiments, in the longitude-time graph 204, a scalar stamp 274 and / or object identifier stamp 282 may be displayed. Similarly, a time value, scalar value, and / or object identifier may be displayed for an identified pixel within the longitude-latitude graph 212. Moreover, a longitude value, latitude value, and / or object identifier may be displayed for an identified or user selected pixel within the scalar-time graph 208. In some cases, the scalar stamp274 and / or object identifier stamp 282 may be displayed near (e.g., within a few pixels of) the point marker 256. The scalar stamp 274 can display a scalar value corresponding to a point associated with the identified (e.g., highlighted) pixel. As shown, the scalar value could be, for example, “12.1 VMag.” Similarly, the object identifier stamp 282 may display an object identifier (e.g., object name) corresponding to the point associated with the identified pixel. As shown, the object identifier could be, for example, 27820: 11003 (AMC-9 (GE-12)). In some embodiments, as noted above, a latitude stamp (not shown) can be displayed. The latitude stamp may be displayed near the point marker 256 and may display a latitude value corresponding to the point associated with the identified pixel.

[0070] In some examples, within the longitude-time graph area 228, the visualization display 300 may include one or more sets of longitude-time points. The one or more sets of longitude-time points may correspond to one or more pixels. Each set of longitude-time points may correspond to data on one or more orbital objects around the planet. Each set of longitudetime points may correspond to a set of identifiers. The set of identifiers may include a longitude value, a latitude value, a time value, a scalar value, and / or an object (e.g., name) identifier. Each set of identifiers may be obtained from one or more photographs. The photographs may contain image data from which one or more identifiers of the set of identifiers can be obtained (e.g., through algorithm).

[0071] A user may be able to adjust the display settings of the visualization display 200 by interacting with the GUI or by a textual command via the chat box. For example, the user may be able to switch a view type, filter what types of points are displayed, turn on or turn off longitudetime map, or change other settings. In some embodiments, a user may be able to pan and zoom within one or more graphs in the visualization display 200. The user may give a panning input and / or a zooming input via the GUI or the text-based Al assistant. The visualization display 300 may be configured to allow simultaneous manipulation of multiple graphs. For example, in response to a user input to pan or zoom the first time axis 220 or the second time axis 244, the system may set the lower-time limit 304 equal to the lower-time limit 504 and / or set the uppertime limit 308 equal to the upper-time limit 508. Similarly, in response to a user input to pan or zoom the first longitude axis 224 or the second longitude axis 236, the system may set the lower-longitude limit 312 equal to the lower-longitude limit 412 and set the upper-longitude limit 316 equal to the upper-longitude limit 416.

[0072] In some examples, the scalar-time graph 208 may show one of a number of possible scalar values. For example, the scalar may refer to a magnitude, such as an intensity of reflected light. However, a number of other scalar values are possible, such as a size, a projected area, a temperature, a mass, a radar cross section, an altitude, an inclination, a delta- V, a time until a certain event, a probability of a certain event, etc.

[0073] As noted above, the system may be configured to store dozens of petabytes of data. This can provide a variety of challenges. One of which is how the data are displayed in a way that is helpful to a human user. Accordingly, in certain embodiments, the visualization display 200 may be configured to divide a graph (e.g., the longitude-time graph 204) into a plurality of pixels. Each pixel may represent a corresponding bin of data. Each bin can be configured to store historical and / or contemporary data as well as metadata.

[0074] In some designs, the system is configured to receive a selection from a user of a target object identifier. In certain embodiments, the system displays metadata stamps for each unique object identifier present in the bin. The visualization display may implement a color scale or gray scale to provide information about the number of unique orbital object identifiers in a bin.

[0075] The system can be configured to identify one or more values (e.g., by various metadata time stamps described herein) associated with a default data set.

[0076] Figure 3 schematically illustrates an example report 300 that may be generated by the system shown in Figure 1 in response to a request for SDA data received from a user via the text-based Al assistant or the GUI of the system shown in Figure 1. In some embodiments, the system main generate the report 300 in response to a user input received from the chat box. In some examples, the user input may comprise a description in natural language indicating two orbital objects (e.g., identified by names or codes), a time window, a request for detecting a maneuver of at least one of the orbital objects within the time window, and certain parameters affected by the maneuver. In some examples, the user input may comprise two orbital objects (e.g., identified by names or codes), and a request for generating a spot report and the system may determine information that should be generated for the spot report. In the example shown, the report 300 may comprise a proximity spot report in response to a maneuver performed by Satellite C. The proximity spot report includes a display portion and text portion including information for two satellites (e.g., satellites C and D). The report shows that Satellite C maneuvered at 17:24:02 on August 12, 2019, with a Delta-V of 1.58 m / s. The apogee increased from GEO-28km to GEO-10km and the perigee decreased from GEO-42km to GEO-55km. Satellite C matched the inclination of Satellite D, and a Delta-V of 1.82 m / s is required to enter proximity operations. The second proximity spot report 1254 indicates that the conjunction is expected to occur with a minimum distance of 16 km + / - 1.2 km at 15:31:32 on August 14, 2019. The maneuver will give Satellite C a solar lighting advantage.

[0077] In some embodiments, the main system 109 shown in Figure 1 may comprise additional analytical tools, visualization displays, and may generate other types of SDA data, as described in U.S. Patent No. 10402672, issued September 3, 2019 and U.S. Patent No. 10976911, issued April 13, 2021 the entire contents of which are incorporated by reference herein in their entirety for providing access and analyzing raw data and generating SDA data buy processing raw data.

[0078] Figure 4A illustrates an example Al enhanced interface 405 for receiving and responding to queries for SDA data received from a user via the corresponding text-based Al assistant, a query received via a text prompt in natural human language, and the resulting display and text generated by the Al enhanced interface 405. In this example, the display comprises a globe and plurality of points distributed around the globe representing orbital objects. In some embodiments, the user may use the text prompt to ask the system to highlight one or more points associated with specific orbital objects. In some embodiments, the display can be interactive and may allow the user to select points using an input device (e.g., a mouse). Once a point is selected the corresponding information may be displayed provided as a textual response in the chat box.

[0079] For example, a user may enter text into the chat interface requesting the system display any predicted conjunctions with orbital object XI for the next week. The system may then use an Al-based system, such as a large language model, to convert the request to an embedding. As used herein, an embedding may refer to a vector representation of input information. The AL based system may then use the generated embedding to identify related embeddings stored within a database. The database may be used to store SDA data as a plurality of embeddings. The SDA data embeddings may be generated by the same, or a different, machine learning model as the model used to implement the chat interface. The related embeddings may, for example, represent a current orbital path of XI, a predicted change in the orbit of XI, a previously-identified conjunction between XI and a second object, an event that may affect the movement or function of XI (e.g., a coronal mass ejection intersecting a predicted location of XI), or any other SDA datathat may be associated with XI. When a related embedding is identified within the database, or another storage location of the system, then the system may determine a preferred mode of display. The system may convert a related embedding to an original format for further use by the system.

[0080] To determine the preferred mode of display, the system may determine the type of information represented by the embedding (e.g., an orbital path, a conjunction time, a current object location, etc.). Based on the type of information represented by the embedding, the system may then select the mode of display from a plurality of potential modes of display. Modes of display may include, but are not limited to, a 3-dimensional image of the Earth and surrounding space, a two-dimensional representation of the Earth where orbital paths are overlaid on the image, a structured text-based object for representing SDA information, a plain text response, or a multimodal response. The mode of display may be determined using a rules-based approach defining types of information associated with different modes of display. Alternatively, a machine learning model may be used to determine a preferred mode of display based on the request, the retrieved embedding information, the type of information represented by the retrieved embedding, a display type of the user providing the request, or any other information available to the system.

[0081] When a mode of display has been determined, the machine learning model may then generate instructions to cause the visualization system 190 to display the information from a retrieved embedding to the user. The visualization may include an image, a video, text, or a sound. The visualization may be multimodal, and the visualization may be interactive. In the present example, the visualization of the Earth may be interactive and allow the user to click and drag the Earth image to rotate the Earth. In response to the Earth being rotated, the visualization system 190 may update the position of at least a portion of the plurality of points to represent an updated relative position of an orbital object. Where the user provides additional information through the text interface, the visualization system 190 may process the additional information to update the presented display. In some embodiments, the visualization system 190 may determine that a second mode of display better represents retrieved information, and may change the mode of display (e.g., from a 3-dimensional image of the Earth to a 2-dimensional image of the Earth with orbital tracks overlaid). Other embodiments are possible.

[0082] Figure 4B illustrates an example user interaction 410 with the Al assistant shown in Figure 4A. In this example, the user has provided a text-based query in natural language as input to the Al assistant, and the Al assistant has generated a response including informationrelevant to the query in text form, but has not caused an update of the visualization. The visualization may not be updated for many reasons. For example, the visualization may not be updated because the requested information is already present in the visualization, the Al assistant has determined that no visualization is appropriate to display information responsive to the query, or because a visualization that would respond to the query is determined not to be useful (e.g., because the view provided would not be helpful for the requesting user to understand an element of the text-based response). In some embodiments, the Al assistant may query the user (e.g., via the chat interface) to determine whether the visualization is to be updated. Advantageously, minimizing updates to the visualization display where such updates are not requested, or determined not to be useful, may reduce overall usage of computing resources associated with updating the visualization display.

[0083] Figure 4C illustrates an example user interaction 415 with the Al assistant resulting in presentation of a selection display. The selection display in this example refers to conjunctions. Selecting a conjunction from the selection display of this example may result in the visualization (shown behind the conjunction display) being updated to visually represent at least a portion of the selected conjunction information. For example, if the user selects one of the conjunction options presented in the selection display, the 3-dimensional visualization behind the selection display may be updated to present a 3-dimensional visualization of the orbit of the orbital object associated with the conjunction, and a current or last known position of the orbital objects. Additionally, the Al assistant has generated a text response to the user query, and the text response is presented to the user within the text-based interface.

[0084] Figure 4D illustrates an example user interaction 420 with the Al assistant resulting in text-based information and visualization information being generated and provided to the user through the user interface. In this example, the Al assistant receives a request for information about the Hubble space telescope following a request for information about Shiyan 12-01. It should be noted that in this example, the Al assistant is shown to be maintaining a history of the text-based chat between the user and the Al assistant that provides context for the portion of the second user request that says “[w]hat about” allowing the Al assistant to understand that the user is requesting information about the Hubble space telescope. Additionally, in response to accessing information for the Hubble space telescope, the Al assistant generates visualizationinformation to cause an updated of the 3-dimensional visualization shown in Figures 4A-4C. The updated visualization may show, for example, an orbital path of the Hubble space telescope.

[0085] Figure 4E illustrates an example of a visualization 425 generated in response to a user request. In this example, the visualization shows an orbital track for two orbital objects overlaid on a two-dimensional image of the Earth.

[0086] Figure 4F-4G illustrate example user interactions 430 and 435 with the textbased Al assistant (e.g., as discussed in relation to Figure 4A) showing queries for SDA data and response generated, in natural human language. For example, Figures 4F-4G illustrate a user interaction where a text-based Al assistant provides SDA information to a user in a text format. The SDA information of this example is provided in natural language in response to a natural language query from the user.

[0087] In another example, Figure 4H illustrates a user interaction 440 where the Al assistant described previously herein provides a text response to a user query in the example chat box provided on the left of this user interface. The Al assistant further causes generation of a visualization associated with the information provided in the text response, the visualization showing a 2-dimensional representation of the orbital path of one or more orbital objects on the right side of this example user interface.

[0088] Figure 5 illustrates an example computing system 500 that may be used to implement various functionality of the SDA analysis system, or portions thereof, described herein.

[0089] In some embodiments, the computing system 500 may be implemented using any of a variety of computing devices, such as server computing devices, desktop computing devices, personal computing devices, mobile computing devices, mainframe computing devices, midrange computing devices, host computing devices, or some combination thereof.

[0090] In some embodiments, the features and services provided by the content fdter system 120 may be implemented as web services consumable via one or more communication networks. In further embodiments, the computing system 500 is provided by one or more virtual machines implemented in a hosted computing environment. The hosted computing environment may include one or more rapidly provisioned and released computing resources, such as computing devices, networking devices, and / or storage devices. A hosted computing environment may also be referred to as a “cloud” computing environment.

[0091] In some embodiments, as shown, a content filter system 120 may include: one or more computer processors 502, such as physical central processing units (“CPUs”); one or more network interfaces 504, such as a network interface cards (“NICs”); one or more computer readable medium drives 506, such as a high density disk (“HDDs”), solid state drives (“SSDs”), flash drives, and / or other persistent non-transitory computer readable media; one or more input / output device interfaces 508; and one or more computer-readable memories 510, such as random access memory (“RAM”) and / or other volatile non-transitory computer readable media.

[0092] The computer-readable memory 510 may include computer program instructions that one or more computer processors 502 execute and / or data that the one or more computer processors 502 use in order to implement one or more embodiments. For example, the computer-readable memory 510 can store an operating system 512 to provide general administration of the computing system 500. As another example, the computer readable memory 510 can store additional functionality 514.

[0093] Figure 6 illustrates example data flows within an example environment for providing automated visualization, identification, tracking, and prediction of space objects according to some embodiments. The example environment includes a machine learning-based response generation and data access system 610, a data collection system 620, a user interface system 630, a data storage system 640, and a system coordination endpoint 650.

[0094] The machine learning-based response generation and data access system 610 is configured to receive requests directed to an Al assistant (e.g., a prompt), and generate a response to the request using a machine learning model. The machine learning-based response generation and data access system 610 includes a database 612, an Al engine 614 (e.g., the Al engine 111 of FIG. 1), and a vector search system 616.

[0095] The database 612 is a database configured to store SDA data in a vector representation. The vector representations may be embeddings generated by a machine learning model (e.g., the Al engine 614, or a machine learning model of the vectorized information generation module 622) and the embeddings may represent SDA data as a multi-dimensional vector. The vector representation may be more efficiently processed by a machine learning model, such as the Al engine 614 described herein.

[0096] The Al engine 614 of the machine learning-based response generation and data access system 610 may be the same as the Al engine / platform described previously in relation toFIG. 1 herein. Alternatively, the machine learning-based response generation and data access system 610 may be a different machine learning system configured to provides responses to requests received from the user interface system 630 via the system coordination endpoint 650. For example, the Al engine 614 may include an LLM, and the LLM may accept input in the form of text. To generate a response, the LLM of this example may communicate with the vector search system 616 (e.g., via an API of the vector search system 616) to access information stored in the database 612. To communicate with the vector search system 616, the Al engine 614 may transform at least a portion of the request received from the user interface system 630 into a vector format (e.g., an embedding) to allow the vector search system 616 to identify relevant information based on a vector comparison. The Al engine 614 of this example may then use the information received from the vector search system 616 when generating a text-based response and visualization information to be transmitted to the user interface system 630 as a response.

[0097] The vector search system 616 is a set of instructions, or computing system, of the machine learning-based response generation and data access system 610 configured to perform a vector search on the vectorized information stored in the database 612. The vector search functionality may be used in part to allow the Al assistant to identify SDA data related to a user query, for example a user query received via the example interfaces in Figures 4A-4D. Identifying SDA data related to the user query may be based in part on the vector search system 616 determining a vector similarity between a vectorized version of a portion of the user query and the SDA data stored in vector form in the database 612.

[0098] The data storage system 640 is a data storage system where SDA data collected from a data collection system is stored. In some embodiments, the data storage system 640 may be a plurality of data storage systems 640. For example, each SDA data collection system (e.g., expert labelled SDA data, automatically generated SDA data from a telescope data collection network, two-line entity data for orbital objects, and the like) may have an associated data storage system 640. The SDA data may be stored in a database implemented on computer storage hardware, such as a hard disk drive, solid state drive, CD, or other storage medium. The data storage system 640 may receive data in real time from some data sources. The data storage system 640 may receive data at regular intervals from additional data sources. Further, the data storage system 640 may receive information from a data source when an alert is generated, and alerts may be generated at irregular intervals. For example, a data source may generate automated, or semi-automated alerts (e.g., human review may be a step in the generation of the alert) related to a potential conjunction, maneuver of an orbital object, or a change in an orbital object’s properties (e.g., an impact, loss of communication, loss of control, and the like). Additionally, an alert may be generated when a manual reviewer has completed tagging at least a portion of an image the includes orbital object information (e.g., an image generated using a telescope) and stored in the data storage system 640. The data storage system 640 may further store an ontology for an orbital object. The ontology may store a history of information related to the orbital object (e.g., maneuver information, conjunction information, battery level information, deployment information, communication status information, and the like). The ontology may be useful for constructing a history of an orbital object, a current orbit of an orbital object (e.g., based on a last known orbit and maneuver information stored in the ontology), or to otherwise determine historical information for an orbital object.

[0099] The data collection system 620 is a computing system configured to request or access SDA data from various SDA data sources (e.g., one or more data storage systems 640). The data collection system 620 may access the SDA data sources via the system coordination endpoint 650, or via a network connection (e.g., a local area network, wide area network, the Internet, and the like). Some information may be accessed or received at regular intervals from a data storage system 640. Additional information may be received from a data storage system 640 when an alert is generated by a data source storing information at the data storage system 640. Receiving an alert from the data storage system 640 when the alert is generated may allow for the machine learningbased response generation and data access system 610 to provide up-to-date, or near real-time, information to a user of the user interface system 630. Providing such near real-time or up-to-date information may allow for improved decision-making by a user of the user interface system 630 by allowing for a decision to be made based on the most current available information.

[0100] The data collection system 620 of this example includes a vectorized information generation module 622. The data collection system 620 may include a machine learning model configured to generate a vector representation (e.g., an embedding) of the accessed or received SDA data for the data collection system 620. The machine learning model may be the same machine learning model implemented by the Al engine 614, or may be a different machine learning model. Alternatively, the vectorized information generation module 622 may be anencoder for a machine learning model, and may be the same or a different encoder than an encoder of the Al engine 614.

[0101] The system coordination endpoint 650 manages or enables communication between the data collection system 620, machine learning-based response generation and data access system 610, user interface system 630, and data storage system 640. For example, the system coordination endpoint 650 may be configured to convert a request received from a first system (e.g., the user interface system 630) to a format accepted by a second system (e.g., the machine learning-based response generation and data access system 610). The format may be based on an API for a given system. The system coordination endpoint 650 may perform additional functions, for example, when a request is sent to the machine learning-based response generation and data access system 610 that exceeds a maximum token length acceptable to the machine learning model of the Al engine 614, the system coordination endpoint 650 may trim tokens (e.g., by removing a portion of the request) to reduce the size of the request to less than the maximum input token size of the machine learning model.

[0102] The user interface system 630 is a system for providing the text-based Al assistant and visualization display described previously herein. The user interface system 630 may implement the visualization system 190 described above herein. The user interface system 630 includes a visualization data augmentation system 632, and an interface module 634. The user interface system 630 may access information from a data storage system 640 to assist in the generation of a visualization to be presented to a user. The accessed information may be based in part on the visualization information received from the machine learning-based response generation and data access system 610 in response to a request. For example, a request may be directed to an upcoming conjunction of an orbital object. The machine learning-based response generation and data access system 610, in the response provided to the user interface system 630, may indicate that the orbital object has a potential upcoming conjunction with a second orbital object, based on an alert received from a data storage system 640. The visualization data augmentation system 632 may then request additional information (e.g., two-line entity information) from the data storage system 640 to generate a three-dimensional display of the orbital paths of the orbital object and the second orbital object with an indicator of the predicted conjunction location and an estimated conjunction time. The interface module 634 is configured to generate the interactive graphical user interface provided to a user of the user interface system630 that is able to display the text-based Al assistant interface and visualization display. The interface module 634 is further configured to receive requests via the interactive graphical user interface and provide the request a relevant system (e.g., the machine learning-based response generation and data access system 610 via the system coordination endpoint 650).

[0103] Figure 7 illustrates an example interaction between a user requesting information from a system for providing automated visualization, identification, tracking, and prediction of space objects according to some embodiments. The user, in this example, transmits a message to the user interface system 630 via the interface module 634. The message may be a natural language request for information associated with an orbital object, or SDA information generally (e.g., predicted conjunctions). The message may be provided as text, an image, audio, or may be multimodal. The message is then forwarded to the visualization data augmentation model 632 at (2) and converted to a token form by the visualization system 190, or by a machine learning provider system (e.g., a third-party provider of a machine learning model that may provide the model, model weights, or other model information to be executed by the visualization system 190 or machine-learning based response generation and data access system 110, or may execute a machine learning model for the visualization system 190). The token form of the message may then be appended to, or otherwise combined with, an available chat history in token form to generate a prompt. The chat history may be stored in token form, or may be converted to token form in response to receiving a user message. The prompt may then be trimmed to reduce the size of the prompt, for example by trimming the tokens at (3). For example, a machine learning model used to provide a response to the message may have an input token limit. In this example, when the prompt, in token form, exceeds the input token limit at least a portion of the prompt may be deleted to avoid exceeding the input token limit of the machine learning model. In some embodiments, the prompt may be trimmed to minimize redundant information (e.g., multiple requests for information related to the same orbital object) to allow for more efficient generation of a response by the machine learning model.

[0104] When the prompt has been generated, the prompt is transmitted to the Al engine 614 (e.g., as a processed chat history) at (4) to cause a machine learning model to generate a response. To generate the response, the machine learning model may perform a vector search by connecting to the vector search system 616 at (5) to cause the vector search system 616 to perform the vector search at (6). A vector search may be performed to identify embeddings that are likelyto be associated with the prompt based on a determined vector similarity between a portion of the prompt in vector form, for example based on a cosine similarity between a stored vector and a vectorized portion of the prompt. It should be understood that while the term vector as used herein may refer to an embedding. The vector form may be generated by the machine learning model, for example when the prompt is received. When the vector search is complete, the Al engine 614 receives relevant results at (7) (e.g., vectorized information having similarity to a vectorized portion of the prompt that satisfies a threshold value). The relevant results in some examples may be a document, portion of a document, unstructured or structured information, or a location of relevant information. Based on the relevant results, the Al engine 614 in this example generates a response to the prompt at (8). Generating the response may include, for example, accessing a document represented in the received vectorized information. The response may include text, images, video, audio, or multimodal information. Additionally, the response may include visualization information useful for generating a visualization responsive to the prompt.

[0105] The Al engine 614 then transmits the response to the visualization system 190, or the visualization data augmentation module 632, at (9). The visualization data augmentation module 632 may forward the message to the interface module 634 at (10) to cause the interface module to display the response at (11). As discussed previously, the visualization system 190 may perform some or all of the functions described in relation to the elements described with respect to FIG. 6. Therefore, the visualization system 190 may display the response, or cause display of the response on a user device (e.g., via an interface shown in FIGS. 4A-4H). For example, the visualization system 190 may provide a text portion of the response in the chat interface and generate a visualization associated with the response and display the visualization in a second portion of the graphical user interface.

[0106] Figure 8 illustrates example interactions between computing systems for retrieving SDA data as part of providing automated visualization, identification, tracking, and prediction of space objects according to some embodiments. As shown, a scheduler (e.g., the system coordination endpoint 650) may communicate with, or be part of, an SDA data server configured to access or store SDA data from a plurality of SDA data sources. The schedule may initiate the access or storage of SDA data at (1) by triggering a poll of the storage locations storing SDA data (e.g., a historical data server 140 or other data storage system 640). The SDA data server (e.g., historical data server 140) may include a visualization data augmentation module 632 thatmay communicate with each SDA data source via the network 144. The SDA data server, or the visualization data augmentation module 632, may connect to each SDA data source and transmit a request for SDA data as shown at (2) and (5). In response, each SDA data source may provide the requested SDA data (e.g., two-line element (TLE) data at (3) and conjunction data at (6)). The returned SDA data may include, for example, two-line element data, orbital object data, conjunction data, and the like. In some embodiments, the SDA data server may additionally generate vector embeddings representing the received SDA data, and store the vector embeddings for later use when performing a vector search during generation of a response to a user query (e.g., the TLE data stored at (4)). Additionally, the request for data from the SDA sever may be based in part on SDA data received from an SDA data source in response to a previous request. The received SDA data used in the request may be received from the same or a different SDA data source. In some cases, the SDA data server (e.g., the visualization data augmentation module 632), may access data from the vector search system 616 that has not been previously stored by the SDA data server, for example data stored by a different SDA data server, or automatically collected or received from a data source by the vector search system 616. For example, the SDA data server may connect to the vector search system 616 at (7) and obtain UCS Satellite Database data at (8). The SDA data server may then augment conjunction data, or other SDA data, at (9) and generate vector embeddings from the augmented data. The vectorized augmented conjunction data may then be provided to the vector search system 616 for later use at (10).

[0107] Figure 9 schematically illustrates an example of communications between computing systems to enable a chat interface for a system providing automated visualization, identification, tracking, and prediction of space objects according to some embodiments. The data storage system 640 (e.g., historical data server 140, or another SDA data server) provides SDA data that may include predicted conjunctions, predicted maneuvers by orbital objects, and combined gravitational action to a database system (e.g., the vector search system 616) for storage. The SDA data may be provided in its original format, a processed format generated by the SDA data sever, or in vector form as an embedding. The database system is in communication with an Al search indexer (e.g., as part of the Al engine 614) that indexes the SDA data stored in the database to allow for more efficient searching in response to a user query through the Al assistant. The Al search indexer is in communication with a machine learning model (e.g., a large language model) allowing the machine learning model to request SDA data from the database through thesearch indexer. In some cases, the machine learning model may be an element of the Al engine 614 that is executed by the Al engine 614. The machine learning model is in communication with the interactive chat interface 111 to receive user requests (e.g., via the interface shown in FIGS.4A-4H) and generate responses to the requests by accessing SDA data through the search indexer.

[0108] While reference is shown to specific computing systems, providers, or software, it should be understood that these systems, providers, and software are exemplary only and that the various functionality described herein may be implemented with any suitable system, provider, or software.Example Embodiments

[0109] Below are a number of nonlimiting example embodiments described above. These examples are for illustration purposes and should not be construed to limit the disclosure above in any way.

[0110] In a 1st Example, a system is disclosed comprising: a computer-readable memory; and one or more processors programmed by executable instructions to at least: receive, via a graphical user interface, a request to provide space domain awareness (SDA) information; transmit the request to a machine learning model system to cause a machine learning model to generate a response, wherein the response comprises text information and visualization information; receiving the response from the machine learning model; generating, based on the visualization information, a visualization responsive to the request; and updating the graphical user interface to display the text information and the visualization.

[0111] In a 2nd Example, the system of Example 1, wherein the graphical user interface comprises a plurality of interface elements, wherein an element of the plurality of interface elements is a text-based chat element, and wherein the text information is displayed in the textbased chat element.

[0112] In a 3rd Example, the system of any of Examples 1-2, wherein a second element of the plurality of interface elements is a visualization element, and wherein the visualization is displayed in the visualization element.

[0113] In a 4th Example, the system of any of Examples 1-3, wherein to generate the response, the machine learning model system is configured to: transmit a request to a searchindexer associated with a vector database, wherein the vector database comprises SDA data converted to a vector format; and receive, from the search indexer, SDA data in the vector format.

[0114] In a 5th Example, the system of any of Examples 1-4 wherein the machine learning model is a large language model.

[0115] In a 6th Example, the system of any of Examples 1-5, wherein the one or more processors are further programmed by executable instructions to at least cause display of the interactive graphical user interface on a user device.

[0116] In a 7th Example, the system of any of Examples 1-6, wherein to generate the response, the machine learning model is further to generate an embedding based on the SDA information associated with the request; transmit the embedding a vector search system to cause the vector search system to identify SDA information relevant to the request; and receive the identified SDA information from the vector search system.

[0117] In an 8th Example, the system of Example 7, wherein the vector search system identifies SDA information relevant to the request based on a comparison between the embedding an a second embedding associated with the identified SDA information satisfying a threshold similarity.

[0118] In a 9th Example, the system of any of Examples 1-8, wherein the graphical user interface comprises a display element comprising a representation of a plurality of orbital objects, and wherein updating the graphical user interface further comprises updating at least one representation of an orbital object of the plurality of orbital objects.

[0119] In a 10th Example, a computer-implemented method is disclosed comprising: receiving, from a user interface system, a request to provide space domain awareness (SDA) information; generating a vector representation of the request; generating a result of a vector search of an SDA database based on the vector representation, wherein the SDA databased comprises a plurality of SDA information items, wherein the SDA information items of the SDA database are stored in a vector form, wherein the vector search is based on determining a vector similarity between the vector representation of the request and the vector format of an SDA information item, and wherein the result of the vector search is at least one SDA information item in vector form having a closest vector similarity to the vector representation; applying the result and the vector representation as input to a machine learning model to cause the machine learning model to generate a natural language response to the request, wherein the response comprises at least aportion of the result, wherein the result further comprises visualization information associated with the at least one SDA information item of the result; and causing display of the response based on transmitting the response to the user interface system, wherein causing display of the response comprises generating a visualization associated with the at least one SDA information item of the result based on the visualization information.

[0120] In an 11th Example, the computer-implemented method of Example 10, wherein the result comprises a natural language description of a conjunction identified from the SDA information.

[0121] In a 12th Example, the computer-implemented method of any of Examples 10- 11, wherein the visualization comprises an indication of a location of at least one orbital object.

[0122] In a 13th Example, the computer-implemented method of any of Examples 10- 12, wherein the request to provide SDA information comprises a natural language request.

[0123] In a 14th Example, the computer-implemented method of any of Examples 10- 13, wherein the request to provide SDA information comprises an identifier associated with an orbital object.

[0124] In a 15th Example, the computer-implemented method of any of Examples 10- 14, wherein the visualization comprises an expected location of the orbital object.

[0125] In a 16th Example, the computer-implemented method of any of Examples 10- 15, wherein the visualization comprises a previous location of the orbital object.Terminology

[0126] All of the methods and tasks described herein may be performed and fully automated by a computer system. The computer system may, in some cases, include multiple distinct computers or computing devices (e.g., physical servers, workstations, storage arrays, cloud computing resources, etc.) that communicate and interoperate over a network to perform the described functions. Each such computing device typically includes a processor (or multiple processors) that executes program instructions or modules stored in a memory or other non-transitory computer-readable storage medium or device (e.g., solid state storage devices, disk drives, etc.). The various functions disclosed herein may be embodied in such program instructions, or may be implemented in application-specific circuitry (e.g., ASICs or FPGAs) of the computer system. Where the computer system includes multiple computing devices, these devices may, but need not, be co-located. The results of the disclosed methods and tasks may bepersistently stored by transforming physical storage devices, such as solid-state memory chips or magnetic disks, into a different state. In some embodiments, the computer system may be a cloudbased computing system whose processing resources are shared by multiple distinct business entities or other users.

[0127] Depending on the embodiment, certain acts, events, or functions of any of the processes or algorithms described herein can be performed in a different sequence, can be added, merged, or left out altogether (e.g., not all described operations or events are necessary for the practice of the algorithm). Moreover, in certain embodiments, operations or events can be performed concurrently, e.g., through multi -threaded processing, interrupt processing, or multiple processors or processor cores or on other parallel architectures, rather than sequentially.

[0128] The various illustrative logical blocks, modules, routines, and algorithm steps described in connection with the embodiments disclosed herein can be implemented as electronic hardware, or combinations of electronic hardware and computer software. To clearly illustrate this interchangeability, various illustrative components, blocks, modules, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented as hardware, or as software that runs on hardware, depends upon the particular application and design conditions imposed on the overall system. The described functionality can be implemented in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the disclosure.

[0129] Moreover, the various illustrative logical blocks and modules described in connection with the embodiments disclosed herein can be implemented or performed by a machine, such as a processor device, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. A processor device can be a microprocessor, but in the alternative, the processor device can be a controller, microcontroller, or state machine, combinations of the same, or the like. A processor device can include electrical circuitry configured to process computer-executable instructions. In another embodiment, a processor device includes an FPGA or other programmable device that performs logic operations without processing computer-executable instructions. A processor device can also be implemented as a combination of computing devices, e.g., a combination of a DSP and a microprocessor, a pluralityof microprocessors, one or more microprocessors in conjunction with a DSP core, or any other such configuration. Although described herein primarily with respect to digital technology, a processor device may also include primarily analog components. For example, some or all of the algorithms described herein may be implemented in analog circuitry or mixed analog and digital circuitry. A computing environment can include any type of computer system, including, but not limited to, a computer system based on a microprocessor, a mainframe computer, a digital signal processor, a portable computing device, a device controller, or a computational engine within an appliance, to name a few.

[0130] The elements of a method, process, routine, or algorithm described in connection with the embodiments disclosed herein can be embodied directly in hardware, in a software module executed by a processor device, or in a combination of the two. A software module can reside in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, a removable disk, a CD-ROM, or any other form of a non-transitory computer-readable storage medium. An exemplary storage medium can be coupled to the processor device such that the processor device can read information from, and write information to, the storage medium. In the alternative, the storage medium can be integral to the processor device. The processor device and the storage medium can reside in an ASIC. The ASIC can reside in a user terminal. In the alternative, the processor device and the storage medium can reside as discrete components in a user terminal.

[0131] Conditional language used herein, such as, among others, "can," "could," "might," "may," “e.g.,” and the like, unless specifically stated otherwise, or otherwise understood within the context as used, is generally intended to convey that certain embodiments include, while other embodiments do not include, certain features, elements and / or steps. Thus, such conditional language is not generally intended to imply that features, elements and / or steps are in any way required for one or more embodiments or that one or more embodiments necessarily include logic for deciding, with or without other input or prompting, whether these features, elements and / or steps are included or are to be performed in any particular embodiment. The terms “comprising,” “including,” “having,” and the like are synonymous and are used inclusively, in an open-ended fashion, and do not exclude additional elements, features, acts, operations, and so forth. Also, the term “or” is used in its inclusive sense (and not in its exclusive sense) so that when used, forexample, to connect a list of elements, the term “or” means one, some, or all of the elements in the list.

[0132] Disjunctive language such as the phrase “at least one of X, Y, Z,” unless specifically stated otherwise, is otherwise understood with the context as used in general to present that an item, term, etc., may be either X, Y, or Z, or any combination thereof (e.g., X, Y, and / or Z). Thus, such disjunctive language is not generally intended to, and should not, imply that certain embodiments require at least one of X, at least one of Y, or at least one of Z to each be present.

[0133] Unless otherwise explicitly stated, articles such as “a” or “an” should generally be interpreted to include one or more described items. Accordingly, phrases such as “a device configured to” are intended to include one or more recited devices. Such one or more recited devices can also be collectively configured to carry out the stated recitations. For example, “a processor configured to carry out recitations A, B and C” can include a first processor configured to carry out recitation A working in conjunction with a second processor configured to carry out recitations B and C.

[0134] While the above detailed description has shown, described, and pointed out novel features as applied to various embodiments, it can be understood that various omissions, substitutions, and changes in the form and details of the devices or algorithms illustrated can be made without departing from the spirit of the disclosure. As can be recognized, certain embodiments described herein can be embodied within a form that does not provide all of the features and benefits set forth herein, as some features can be used or practiced separately from others. The scope of certain embodiments disclosed herein is indicated by the appended claims rather than by the foregoing description. All changes which come within the meaning and range of equivalency of the claims are to be embraced within their scope.

Claims

CLAIMSWHAT IS CLAIMED IS:

1. A system comprising:a computer-readable memory; andone or more processors programmed by executable instructions to at least:receive, via a graphical user interface, a request to provide space domain awareness (SDA) information;transmit the request to a machine learning model system to cause a machine learning model to generate a response, wherein the response comprises text information and visualization information;receiving the response from the machine learning model;generating, based on the visualization information, a visualization responsive to the request; andupdating the graphical user interface to display the text information and the visualization.

2. The system of claim 1, wherein the graphical user interface comprises a plurality of interface elements, wherein an element of the plurality of interface elements is a text-based chat element, and wherein the text information is displayed in the text-based chat element.

3. The system of claim 2, wherein a second element of the plurality of interface elements is a visualization element, and wherein the visualization is displayed in the visualization element.

4. The system of claim 1, wherein to generate the response, the machine learning model system is configured to:transmit a request to a search indexer associated with a vector database, wherein the vector database comprises SDA data converted to a vector format; andreceive, from the search indexer, SDA data in the vector format.

5. The system of claim 1, wherein the machine learning model is a large language model.

6. The system of claim 1, wherein the one or more processors are further programmed by executable instructions to at least cause display of the interactive graphical user interface on a user device.

7. The system of claim 1, wherein to generate the response, the machine learning model is further to:generate an embedding based on the SDA information associated with the request; transmit the embedding a vector search system to cause the vector search system to identify SDA information relevant to the request; andreceive the identified SDA information from the vector search system.

8. The system of claim 7, wherein the vector search system identifies SDA information relevant to the request based on a comparison between the embedding an a second embedding associated with the identified SDA information satisfying a threshold similarity.

9. The system of claim 1, wherein the graphical user interface comprises a display element comprising a representation of a plurality of orbital objects, and wherein updating the graphical user interface further comprises updating at least one representation of an orbital object of the plurality of orbital objects.

10. A computer-implemented method comprising:receiving, from a user interface system, a request to provide space domain awareness (SDA) information;generating a vector representation of the request;generating a result of a vector search of an SDA database based on the vector representation, wherein the SDA databased comprises a plurality of SDA information items, wherein the SDA information items of the SDA database are stored in a vector form, wherein the vector search is based on determining a vector similarity between the vector representation of the request and the vector format of an SDA information item, and wherein the result of the vector search is at least one SDA information item in vector form having a closest vector similarity to the vector representation;applying the result and the vector representation as input to a machine learning model to cause the machine learning model to generate a natural language response to the request, wherein the response comprises at least a portion of the result, wherein the result further comprises visualization information associated with the at least one SDA information item of the result; andcausing display of the response based on transmitting the response to the user interface system, wherein causing display of the response comprises generating a visualization associated with the at least one SDA information item of the result based on the visualization information.

11. The computer-implemented method of claim 10, wherein the result comprises a natural language description of a conjunction identified from the SDA information.

12. The computer-implemented method of claim 10, wherein the visualization comprises an indication of a location of at least one orbital object.

13. The computer-implemented method of claim 10, wherein the request to provide SDA information comprises a natural language request.

14. The computer-implemented method of claim 10, wherein the request to provide SDA information comprises an identifier associated with an orbital object.

15. The computer-implemented method of claim 10, wherein the visualization comprises an expected location of the orbital object.

16. The computer-implemented method of claim 10, wherein the visualization comprises a previous location of the orbital object.

Images