Onboard AI Cloud Detection with PUS for Satellite Power Savings

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing onboard cloud detection systems for complex space missions face challenges such as hardware incompatibility with space-grade standards, high costs, increased complexity due to parallel AI processing, and excessive power consumption, which are not optimized for efficient resource management.

Innovation Solution

A dedicated onboard cloud detection system utilizing a packet utilization standard (PUS) that includes a camera, data processing unit, and AI algorithms to detect clouds, enabling or disabling main payload equipment based on cloud coverage, thereby optimizing memory and power usage.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If a dedicated AI image processing unit is implemented for each instrument or a centralized DPU processes multiple data streams in parallel, then cloud detection capability is improved, but device complexity increases

Engineering Contradiction:
Improvecloud detection capabilityVSAvoidhardware complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent implements a centralized DPU that serves multiple instruments simultaneously, making the cloud detection system universal rather than dedicated to each payload. The DPU processes data streams from multiple sources through a unified architecture, reducing overall system complexity while maintaining detection capabilities across all instruments.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

The patent merges cloud detection functionality into the existing DPU architecture rather than adding separate processing units. By combining cloud detection with the centralized data processing function, the system avoids parallel hardware implementations and reduces device complexity while maintaining reliable cloud detection across multiple instruments.

Inventive Principle:
Principle #5Merging (Combining)

2Reliability

If the payload operates continuously to ensure operational readiness, then detection capability is improved, but power consumption increases

Engineering Contradiction:
Improvepayload operational readinessVSAvoidpower consumption
Core Design Contradiction:
ReliabilityVSUse of energy by moving object

Solution Approach 1:

The patent implements periodic cloud detection at predetermined intervals rather than continuous payload operation. The DPU acquires images from the camera at specific time intervals and processes them for cloud detection, allowing the payload to enter low-power states between detection cycles while maintaining operational readiness when needed.

Inventive Principle:
Principle #19Periodic action

Solution Approach 2:

The system uses a dedicated camera and DPU for cloud detection that operates independently from the main payload processing. This self-service detection capability monitors cloud conditions without requiring the main payload to remain fully operational, enabling power savings while maintaining detection functionality.

Inventive Principle:
Principle #25Self-service

3Loss of information

If cloud coverage data is continuously stored on-board for analysis, then data availability is improved, but memory space consumption increases

Engineering Contradiction:
Improvecloud data availabilityVSAvoidon-board memory space
Core Design Contradiction:
Loss of informationVSQuantity of substance

Solution Approach 1:

The patent extracts only the essential cloud detection results (cloud presence/absence at detection intervals) rather than storing all raw images and processing data. The DPU processes images and stores only the relevant cloud coverage information, significantly reducing memory space requirements while maintaining data availability for mission analysis.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The system stores cloud detection data with varying detail levels based on local requirements. Full-resolution processing occurs only when needed, while routine cloud monitoring uses compressed or summarized data storage, optimizing memory usage according to the specific quality requirements of different mission phases.

Inventive Principle:
Principle #3Local quality

Data Source

PatentEP4089573B1Onboard ai-based cloud detection system for enhanced satellite autonomy using pus
Publication Date: 2025.08.20 AIRBUS DEFENCE & SPACE GMBH
  • EP4089573B1 patent drawingFigure 1~2
  • EP4089573B1 patent drawingFigure 3
  • EP4089573B1 patent drawingFigure 4

AI summary

An onboard cloud detection system comprising: a camera (1000) configured to acquire images of the Earth at predetermined acquisition intervals; and a data processing unit (2000) comprising: a cloud detection unit (2210) configured to use artificial intelligence, AI, algorithms to detect clouds; a packet utilization standard, PUS, application layer (2230) configured to issue telemetry and/or telecommands corresponding to a predetermined parameter of the output of the cloud detection unit (2210); and an interface configured to distribute the telemetry and/or telecommands to an external hardware and/or an external software terminal (3000, 4000).