Spatial data management system for providing traceability, privacy and quality.

Abstract

The invention relates to a method for managing spatial data in an environment for space and Earth monitoring or orbital operations security, the method being characterised by the following steps: A. Structuring the spatial data into a directed acyclic graph (DAG); B. Associating a creation date and traceability metadata with each node, making it possible to link each node to the parent nodes from which it is derived; C. Including, for each node, digital quality indicators and a confidentiality level specific to the data that it contains; D. Transferring the metadata containing the transmitted indicators and levels from each node to the derived nodes; and E. Managing and monitoring the data transformations according to the quality and confidentiality metadata.

Description

A spatial data management system that ensures traceability, confidentiality, and quality.

[0001] The present invention relates to space data management applied to space surveillance, surveillance from space, and the security of orbital operations. It specifically concerns the tracking and characterization of objects in orbit, encompassing the collection, storage, and aggregation of multi-source data (such as data from telescopes, radars, lasers, etc.), the transformation, analysis, and distribution of data from sensors such as images, astrometric and photometric measurements, and finally, higher-level data produced from measurements such as orbits or predictions of dangerous close approaches.

[0002] The quality of space data and its processing is crucial for space surveillance for the purpose of tracking space objects and preventing collisions, producing orbital trajectories, managing satellite fleets, ensuring the reliability of warning systems essential for protecting satellites and manned missions, and real-time tracking applications. State of the art

[0003] Prior art patent EP3482543 describes an observation system comprising at least one observation device, each including at least one observation unit configured to observe at least one observation object according to at least one first instruction data set. The observation system includes at least one access control device configured to control access to the observation device by at least one access entity. The control organization includes at least one first peer-to-peer module assigned to the observation device. The first peer-to-peer module is configured to transmit at least one first instruction data set to the observation unit. The control organization includes at least one peer-to-peer application of at least one peer-to-peer network.The peer-to-peer application is configured to control access to the observation unit by allowing an accessing entity to provide at least one set of instruction data to the first peer-to-peer module, specifically through a second peer-to-peer module of the accessing entity, where the peer-to-peer application is configured to generate at least one observation access transaction agreement concerning an action of accessing the observation device by the accessing entity, and where the peer-to-peer application is configured to store the generated observation access transaction agreement.

[0004] Patent CN111241038B is also known, describing a method and system for real-time satellite data processing designed to handle massive volumes of data generated by satellites. This prior art solution uses the Storm™ distributed computing framework to process data continuously with low latency, high throughput, and scalable processing capacity. It offers the following steps: streaming binary data from satellites to generate transmission frames; parallel classification of transmission frames and source data units, allowing each type of data packet to be processed in parallel; and extraction and generation of data product files without intermediate storage, thus reducing processing times.

[0005] The system uses several distributed computing nodes, each responsible for a specific step, such as frame synchronization, virtual channel separation, and data product generation. A state coordination layer manages the state of the tasks and recovers the nodes in case of failure.

[0006] Chinese patent CN112152699B describes an alternative solution for real-time processing of satellite data, and for combining, classifying, and distributing the data to payload terminals via an optical network. The proposed system includes a processing and distribution system that performs data format analysis, combination, storage, reading, classification, and real-time distribution of the classified data to payload terminals via an optical network.

[0007] Chinese patent CN111241038B proposes an alternative real-time satellite data processing system capable of efficiently and scalably handling large volumes of data by leveraging distributed streaming for improved operational performance. This system builds upon the Storm™ real-time distributed computing framework designed to process large volumes of continuously streaming data with low latency and high throughput, and includes control, worker, and state coordination nodes. The components process data in parallel based on Virtual Channel Identifiers (VCIDs) and Application Process Identifiers (APIDs), enabling scalability and high reliability. Disadvantages of prior art

[0008] Prior art solutions, and particularly the system described in patent CN111241038B, process spatial data in parallel and distributed, but lack a mechanism to precisely track the origin and transformation steps of each data point. Specifically, this system does not include metadata or mechanisms to control data quality based on specific criteria or to enforce confidentiality rules during data transformation and sharing.

[0009] This raises the issue of the lack of data traceability during each transformation and makes it impossible to manage data reliability and confidentiality, or to adapt a data management policy according to the source and quality. Solution provided by the invention

[0010] To overcome these drawbacks, the present invention relates to a method for managing spatial data in an environment of Earth-from-space surveillance or orbital operations security. The method is characterized by the following steps: Structuring the spatial data in the form of a directed acyclic graph (DAG), each node of said directed acyclic graph (DAG) representing a unique unit of spatial data, including, but not limited to, images produced by sensors, measurement data extracted from images or radar frames, or position and velocity vectors, each node being identified by a unique identifier and associated with metadata characterizing, in particular, the type of data, the origin (e.g., the sensor source), the timestamp, each edge representing a transformation between two data nodes or a process applied to the data of a source node to generate a target node.The transformation is defined by a unique digital process identifier, and the metadata associated with the edges includes parameter and version information to ensure traceability of operations performed between the initial node and the derived node. Each node is associated with a creation date and traceability metadata that links each node to the parent nodes from which it is derived. For each node, it includes digital quality indicators and a confidentiality level specific to the data it contains. The metadata containing these indicators and levels is transferred from each node to the derived nodes. Data transformations are managed and monitored based on this quality and confidentiality metadata.

[0011] Advantageously, the process further includes episodic processing steps of the numerical data recorded in the metadata of each node to calculate numerical anomaly data and to record in the metadata of each node the detected anomaly information, to automatically report inconsistencies in the initial data or in successive transformations.

[0012] According to another variant, the process includes a data exclusion processing step based on a compliance criterion of the digital indicator of metadata quality.

[0013] The invention also relates to a spatial data processing system based on a directed acyclic graph (DAG) structure, for implementing a spatial data management method, having the aforementioned characteristics A to E, characterized in that it comprises:

[0014] a) A node creation module structuring spatial data into nodes and edges in a DAG;

[0015] b) A metadata calculation module configured to associate traceability information, quality indicators and confidentiality levels with each node;

[0016] c) A transformation monitoring module configured to track the evolution of data in the DAG according to associated quality and confidentiality indicators;

[0017] d) An automatic alert module that signals anomalies when quality indicators do not meet predefined criteria or when confidentiality issues are detected.

[0018] Advantageously, the node creation module is configured to group multi-source spatial data from various sensors and organize it in the DAG in such a way as to ensure no redundancy and consistent aggregation.

[0019] According to one variant, the monitoring module is configured to block data transmission to end users if the confidentiality level or the node quality indicator is below a predefined threshold.

[0020] Detailed description of a non-limiting example of implementation

[0021] The present invention will be better understood upon reading the following description, concerning a non-limiting example of an embodiment illustrated by the accompanying drawings where:

[0022] The figure represents the table of different types of spatial monitoring data

[0023] The figure represents the tree structure created with the data management system according to the invention. Principle of the invention

[0024] The principle of the invention is based on the digital structure of multi-source spatial data on the one hand, and on the structure of interactions between spatial data on the other. In the context of the invention, a directed acyclic graph (DAG) structure is a mathematical representation of data where the relationships between different types of data are organized in such a way as to avoid loops; that is, it is not possible to return to a starting node by following the edges of the graph. In other words, the graph is directed (each connection between nodes has a direction) and acyclic (without cycles or loops).

[0025] A directed acyclic graph (DAG) structure is a graphical representation composed of directed nodes and edges, where each edge indicates a direction of processing or transformation from a source node to a target node, and where acyclicity guarantees the absence of loops. This structure is used to determine the sequential processes of propagating multi-source spatial data hierarchically, ensuring that each derived element depends on previous elements and propagates information about the confidence level of the original and aggregated data.

[0026] A directed acyclic graph (DAG) structure preserves data immutability. The data processing flow cannot flow backward from a downstream data source to an upstream data source from which it inherits, thus preserving the data and ensuring traceability and authenticity. Spatial data format according to the invention

[0027] The spatial data format according to the invention is designed to meet the complex needs of acquiring, processing, and interpreting collected information related to space activities. Space surveillance data, for example, are of various types, as shown in the figure.

[0028] Nodes are created by associating spatial data belonging to a data level with digital metadata, which includes: a unique UUID, a creation date, a list of the parent node, an algorithm and its parameters, a confidentiality level, a quality rating, and an indicator for detecting anomalies if a metric exceeds a certain threshold. This applies to a spatial monitoring processing chain.

[0029] To deliver data from a space monitoring sensor in the format defined by the invention, several electronic and computer-based means are implemented to ensure the production, conversion, processing, structuring, and transmission of this data. The data originates from radar, optical, or laser sensors. This data can be transformed into formats such as images that allow for the detection of an object's position in pixels, or a radar frame that enables a Doppler measurement.

[0030] A server or computing unit directly connected to the sensor records the raw data and, if necessary, processes the data produced by the sensor to extract astrometric or photometric measurements. Following this recording, a certificate is issued to authenticate the raw data produced, as well as any contextual data (sensor that produced the data, timestamp, etc.).

[0031] Supporting instruments such as a weather station or a precise time-stamping system can enrich the measurement data with environmental data (temperature, cloud cover, acquisition date).

[0032] The images or associated measurements and metadata can be sent to a processing platform for distribution to a third party, for use in catalog building or maintenance, or for use in various services to ensure the safety of space activities. If the station is temporarily disconnected from the central platform, the data produced is buffered until the connection is restored.

[0033] The data and metadata are collected by a centralized platform server, which organizes the data into a directed acyclic graph (DAG). The data and metadata form the nodes and edges, and perform transformations and calculations.

[0034] Consecutive measurements taken during the passage of a single observed space object over the station that acquired the data form a track used by the algorithm and are linked to the estimated orbital state. Furthermore, each track has a rating, an indicator for detected anomalies, and a confidentiality level. All these properties are passed to the orbital state vector, which can be verified by unfolding the data decomposition into a tree structure, as shown in the figure. The state vector was generated from several tracks, themselves generated from measurements, which were in turn generated from images or radar frames.

[0035] Next, an operator is able to apply a data management policy, excluding a lead if the rating is below a quality threshold, if the level of confidentiality excludes its use, or if an anomaly has been detected in the data production chain.

[0036] Application to an Earth observation processing chain

[0037] The principles outlined for space object monitoring data can be applied to Earth observation data. This data originates from onboard sensors, including cameras, spectrometers, temperature sensors, and position sensors, integrated onto a satellite or space platform. These sensors collect raw data in the form of analog or digital signals.

[0038] A microcontroller or embedded processors, particularly those integrated into the sensor system or a nearby computing unit, perform basic data operations such as noise filtering, scaling, and calibration. For large datasets (e.g., images), lossless (or lossy) compression algorithms can be applied to reduce the amount of data transmitted. These transformations are also recorded in the metadata.

[0039] A synchronization and metadata calculation module associates synchronization information and metadata, such as the exact acquisition time and satellite position, with the data.

[0040] An FPGA (Field-Programmable Gate Array) circuit can be configured to structure data into packets, applying the defined data format, for example by organizing the data into nodes of type image, telemetry, etc., and adding headers conforming to the intended format.

[0041] A DSP processor can execute data processing and encapsulation algorithms, transforming the data according to formatting specifications and applying any necessary corrections or conversions. The data can be encapsulated into packets conforming to the CCSDS (Consultative Committee for Space Data Systems) standard, which is commonly used for transmitting space data. This standard ensures compatibility with other space systems.

[0042] Formatted data is temporarily stored in flash memory, allowing for buffering before transmission. Queues store data awaiting processing or transmission, ensuring that each piece of data is sent in the correct order.

[0043] The formatted spatial data is then transmitted by an RF transmission module, which sends the data sequentially to the ground station. This unit must be compatible with the data formats and include integrity protocols (error control, cyclic redundancy check). In some cases, satellites use laser links to transmit the data at high speed to relays or ground stations.

[0044] A ground station receives data from the satellite via antennas and uses receivers to demodulate and decode the signals. Once the data is received, servers perform further processing, if necessary, by applying final steps of formatting, structuring, and organizing it into a directed acyclic graph (DAG). They structure each node (image, measurement, event, etc.) and edge (transformations, calculations) within the graph.

[0045] Data Management and Distribution Platforms

[0046] A relational or NoSQL database can be used to store data in a structured data aggregation (DAG), indexing nodes (data types) and edges (transformations). Using a data lake coupled with a relational database is also possible. Application programming interfaces (APIs) provide access to data in a structured format. These APIs are designed to deliver data in the required format to end users (analytical centers, scientists, specific applications).

[0047] In summary, the electronic and computer process for delivering formatted spatial data in the context of the invention includes sensors, embedded processing units (computing server or FPGA and DSP component), communication modules, and ground servers, all configured to ensure that the data is processed, structured, and distributed according to the directed acyclic graph format, thus guaranteeing the traceability and integrity of the data and their transformations from acquisition to final delivery. Database structure

[0048] The spatial database is structured to include different sections in the spatial data headers. Several metadata groups exist, linked by the unique identifier of the node. Several types of metadata must be stored in the spatial database: Contextual metadata includes the identification of a source (a sensor, its location, its status, a data provider), environmental measurements (temperature, cloud cover, etc.), or information about the data provider (confidence level, average data quality, etc.). The metadata essential for creating the directed acyclic diagram are: The unique identifier of the data or transformation. The timestamp (precise date and time with time zone) of the data creation. The data type (image, measurement, state vector, etc.). The algorithm version and the processing algorithm parameters.The list of parent nodes. The quality and confidentiality metadata specific to each data type. Business data is stored in the spatial database and includes images, measurements, orbits, maneuvers, etc., as described in the.

[0049] In the context of space surveillance, the database records the following primary information: Sensor images: Images are generally stored in grayscale or color with intensity in different spectral bands. Astrometric measurements: Measurements are generally stored as radial distance and / or angular coordinates and / or their derivatives in a topocentric reference frame. Orbital parameters: Defining the dynamic state, orbital parameters are equivalent to position and velocity coordinates. Events: The description of an event such as a close approach, a maneuver, or a behavior. Transformation and Processing Data

[0050] Spatial data and their metadata are used by the various processing workflows according to rules defined based on usage.

[0051] Spatial data can be extracted in CCSDS format with comments containing metadata including all the information necessary for data traceability.

[0052] The transformations applied correspond to the processes applied to the data (e.g., geometric correction, spectral filtering, associating a measurement with an object in a catalogue, determining the orbit, etc.) and their parameters.

[0053] Each treatment is time-stamped and includes the treatment parameters that are associated with the unique identifier of the treatment. Quality and Trusted Metadata

[0054] The metadata includes quality indicators, consisting of qualitative data measurements (e.g., signal / noise ratio, root mean square value after processing, deviation from another source, etc.) and environmental measurements (e.g., a reliability index calculated based on sensor performance such as entropy level, atmospheric conditions, or cloud cover percentage).

[0055] Data can be associated with confidentiality or classification levels to regulate its access and distribution. Some data may require access rights for viewing. Relationships between existing nodes

[0056] The relationship between nodes is established through data transformation algorithms. For example, image processing, including object detection and astrometric reduction, consumes images and returns measurements. Therefore, a measurement is always linked to an image. Image processing is parameterized according to very specific criteria to manage noise and the type of object to be detected. These parameters are stored in the metadata.

[0057] Throughout the chain, the processes and their parameters create relationships between data types. These relationships can exist between two data levels, for example a measurement and a track, or a track and an orbital state vector, but also at the same level, for example several orbital state vectors can give a new orbital state vector.

[0058] In the case where a transformation is replayed from the same data, this will result in a new node which differs by a different processing date, but a node already created cannot be modified. Applications of the invention

[0059] The aim of the invention is the optimization of space data management for applications of space or Earth surveillance, space operations security, and management of satellites and orbital debris.

[0060] Thanks to its directed acyclic graph (DAG) structure, it allows for structured organization and detailed monitoring of spatial data, as well as rigorous management of quality and confidentiality.

[0061] The invention relates, for example, to improving data traceability and history: By structuring data into nodes and links (edges) in a DAG, each transformation and data source is traceable. This makes it possible to trace information back to its origin, verify transformation steps, and understand how data was produced. Since it is not possible to modify a node upstream of a piece of data, the information is guaranteed. This traceability is crucial for validating the accuracy of information, especially in critical contexts such as collision alerts.

[0062] This allows for the efficient management of space operations, such as collision risk management: By consolidating observation data (like tracks) using quality metadata and an anomaly indicator, the invention improves confidence in conjunction detection, that is, the moment when two orbiting objects are closest to a potential collision risk. The resulting collision alerts allow satellite operators to plan avoidance maneuvers and reduce the risk of collision, which is increasingly necessary with the growing amount of space debris and space traffic in general.

[0063] The method according to the invention enables data quality and confidentiality control: By integrating quality and confidentiality metadata into each node of the DAG, this invention ensures that the data used is reliable and that sensitive information is protected. Data quality is inherited from one node to another, thus guaranteeing that only data conforming to quality requirements is used for operational decisions.

[0064] The process also offers support for managing large amounts of multi-source data: The DAG structure allows for the aggregation of data from multiple sensors and sources, while maintaining an organized and structured history of transformations. This simplifies continuous data management, facilitating the merging of diverse information (such as measurements and tracks from different sensors) to obtain a complete and consistent view of the spatial situation.

[0065] Another application context concerns the optimization of catalog construction and maintenance operations, the product of which is used for various event detection and characterization services: Thanks to the precise management of spatial data, including tracks, orbits, and trajectory predictions, the invention helps to plan and adapt the required level of accuracy by selecting only reliable data (based on its quality and consistency, identifying any potential anomalies). Thus, the presence of an anomaly can lead to specific catalog management decisions, such as enhanced object tracking by a sensor.

[0066] Another area of ​​application for the invention concerns the coordination and sharing of data for international collaborations: The standardized DAG structure facilitates the sharing of space data between different operators, space agencies, and international organizations, while ensuring compliance with confidentiality and security policies. This promotes effective cooperation for debris monitoring and the safety of space operations, and fosters trust in the use of the data.

Claims

A method for managing spatial data in a space and Earth surveillance or orbital operations security environment, the method being characterized by the following steps: Structuring the spatial data in the form of a directed acyclic graph (DAG), each node of said directed acyclic graph (DAG) representing a unique unit of spatial data, including, but not limited to, measurement data from specific sensors, and in particular radar data or image data produced by ground or airborne stations, position and velocity vectors, or measurement data from specific sensors, each node being identified by a unique identifier and associated with metadata characterizing the data type, origin (e.g., sensor source), timestamp,and other relevant parameters; each edge representing a transformation between two data nodes or a process applied to the data of a source node to generate a target node, said transformation being defined by a unique digital process identifier; and metadata associated with the edges including traceability and version information, for the traceability of operations performed between the initial node and the derived node. Associate with each node a creation date and traceability metadata allowing each node to be linked to the parent nodes from which it is derived. Include for each node digital quality indicators and a confidentiality level specific to the data it contains. Transfer said metadata containing said indicators and levels from each node to the derived nodes. Manage and monitor data transformations based on said quality and confidentiality metadata. Method according to claim 1 characterized in that it further comprises episodic processing steps of the digital data recorded in said metadata of each node to calculate digital anomaly data and to record in the metadata of each node the detected anomaly information, to automatically report inconsistencies in the initial data or in successive transformations. A method according to claim 1, characterized in that it comprises a data exclusion processing step based on a conformity criterion of the digital quality indicator of the metadata. A spatial data processing system based on a directed acyclic graph (DAG) structure, for implementing a spatial data management method, having features A to E of claim 1 characterized in that it comprises: a) A node creation module structuring spatial data into nodes and edges in a DAG; b) A metadata calculation module configured to associate traceability information, quality indicators and confidentiality levels with each node; c) A transformation monitoring module configured to track the evolution of data in the DAG according to the associated quality and confidentiality indicators; d) An automatic alert module signaling anomalies when quality indicators do not meet predefined criteria or when confidentiality problems are detected. System according to claim 4, wherein the node creation module is configured to group multi-source spatial data, from various sensors, and organize them in the DAG so as to ensure the absence of redundancy and consistent aggregation. System according to claim 4 or 5, wherein the monitoring module is configured to block the transmission of data to end users if the confidentiality level or the node quality indicator is below a predefined threshold. A spatial data processing system based on a directed acyclic graph (DAG) structure, characterized in that it comprises: a) A node creation module structuring spatial data into nodes and edges in a DAG; b) A metadata calculation module configured to associate traceability information, quality indicators and confidentiality levels with each node; c) A transformation monitoring module configured to track the evolution of data in the DAG according to the associated quality and confidentiality indicators; d) An automatic alerting module signaling anomalies when quality indicators do not meet predefined criteria or when confidentiality problems are detected. System according to claim 4, wherein the node creation module is configured to group multi-source spatial data, from various sensors, and organize them in the DAG so as to ensure the absence of redundancy and consistent aggregation. System according to claim 4, wherein the monitoring module is configured to block the transmission of data to end users if the confidentiality level or the node quality indicator is below a predefined threshold.

Images