Multispectral Sensor Fusion for Autonomous Flight Navigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional sensor systems for aircraft, designed for human pilots, face limitations in dynamic range and processing power, which can hinder object identification and navigation, especially in challenging environments like fog or sunlit conditions, due to constraints on color spectrum and dynamic range, leading to potential inability to distinguish objects.
Innovation Solution
A multispectral sensor suite onboard autonomous aircraft, comprising a vision system, RF RADAR, LIDAR, and a processor with a database, capable of receiving and processing sensor data at various frame rates and latencies, and dynamic ranges suitable for flight control and navigation, allowing for object identification and trajectory determination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If sensor data is processed at high frame rates and latencies suitable for flight control, then flight control performance is improved, but processing power and computational resources are consumed
Solution Approach 1:
The patent segments sensor data processing into multiple streams with different frame rates and latencies tailored to specific functions (flight control vs. positioning). This allows high-speed processing only for critical flight control parameters while using lower-speed processing for less time-sensitive positioning data, thereby reducing overall computational resource consumption.
Solution Approach 2:
The patent applies partial processing by selectively processing sensor data at different quality levels and frame rates based on the specific application needs. Not all sensor data requires maximum processing speed - only the portions critical for flight control do, while positioning can tolerate lower processing rates, optimizing the balance between performance and resource usage.
2Ease of operation
If sensor data is processed with limited dynamic range for human pilot consumption, then display compatibility is improved, but object identification capability deteriorates
Solution Approach 1:
The patent segments sensor data processing into multiple output streams: one optimized for human pilot consumption with limited dynamic range and display formatting, and another preserving full dynamic range and spectral information for autonomous object identification algorithms. This allows both human readability and machine precision to coexist without compromising either.
Solution Approach 2:
The patent introduces an intermediary processing layer that converts full-dynamic-range sensor data into human-friendly display formats while simultaneously extracting and preserving critical object identification features. This intermediary layer acts as a mediator that translates between the full-information sensor output and the limited-capability human display system, preventing information loss for autonomous processing.
3Device complexity
If a single video stream is provided for pilot selection, then system complexity is reduced, but adaptability to different environments deteriorates
Solution Approach 1:
The patent implements dynamic sensor selection and data stream generation where the system automatically adapts to different environmental conditions (weather, lighting, time of day) and autonomously selects the most appropriate sensor types and spectral ranges. This dynamic adaptation provides environmental versatility without requiring manual pilot intervention or complex configuration, maintaining low system complexity while achieving high adaptability.
Solution Approach 2:
The patent incorporates feedback mechanisms where sensor data quality and environmental conditions are continuously monitored, and the system automatically adjusts which sensors are active and how data is processed. This feedback loop enables the system to adapt to changing environments in real-time, providing versatility without requiring multiple manually-configurable video streams or pilot expertise in sensor selection.
4Reliability
If transparent display is used for immediate comparison of sensor imagery to real world, then pilot situational awareness is improved, but dynamic range and processing capability are constrained
Solution Approach 1:
The patent segments the display system into transparent overlays for pilot situational awareness and separate processing channels for autonomous analysis. The transparent display presents simplified, human-readable sensor information overlaid on the real-world view, while the full-resolution, full-dynamic-range sensor data continues to be processed independently for autonomous object identification and navigation, eliminating the constraint that transparent displays impose on processing capability.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables precise object identification and navigation in diverse conditions by processing high-dynamic-range imagery without loss of detail, improving the autonomous aircraft's ability to discern objects and maintain accurate flight paths.
Implementation Method 1
a laser imaging detection and ranging (LIDAR) system
Implementation Method 2
a radio frequency (RF) radio detection and ranging (RADAR) system
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A system and method are disclosed for design of a suite of multispectral (MS) sensors and processing of enhanced data streams produced by the sensors for autonomous aircraft (120) flight. The suite of MS sensors is specifically configured to produce data streams for processing by an autonomous aircraft object identification and positioning system processor. Multiple, diverse MS sensors image naturally occurring, or artificial features (towers buildings etc.) and produce data streams containing details which are routinely processed by the object identification and positioning system yet would be unrecognizable to a human pilot. The object identification and positioning system correlates MS sensor output with a-priori information stored onboard to determine position and trajectory of the autonomous aircraft (120). Once position and trajectory are known, the object identification and positioning system sends the data to the autonomous aircraft flight management system for autopilot control of the autonomous aircraft.