Flight Plan Safety Monitor Using 4D Occupancy Overlap
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional offline verification methods struggle to certify autonomous systems due to the complexity of algorithms like AI and ML, hindering the introduction of non-standard algorithms in safety-critical applications such as UAVs, despite their potential benefits.
Innovation Solution
A safety monitor that verifies flight plans by generating four-dimensional occupancy regions for aircraft and hazards, accommodating uncertainties, and determining overlap to ensure safety, allowing non-standard algorithms to generate flight plans independently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional flight planning algorithms are used, then the flight plan can be certified as safe, but the system lacks flexibility to use non-standard algorithms like AI and ML
Solution Approach 1:
The system separates flight plan generation from safety verification into independent modules. The flight plan generator can use any algorithm (including non-standard AI/ML), while the safety monitor independently verifies the plan against hazard data, allowing algorithm flexibility without compromising certification
Solution Approach 2:
A safety monitor acts as an intermediary between the flight plan generator and the execution system. It receives flight plans from any source, verifies them against hazard occupancy regions, and only allows execution if safe, thus enabling nonstandard algorithms while maintaining safety certification
2Productivity
If traditional offline verification methods are used, then safety can be verified, but the verification process is too slow and complex for modern autonomous systems
Solution Approach 1:
Hazard occupancy regions are pre-computed and stored before flight plan verification. This preliminary preparation of hazard data enables rapid verification of flight plans without complex real-time calculations, increasing verification speed while maintaining safety assurance
Solution Approach 2:
The system uses simplified geometric representations (occupancy regions) as copies of complex hazard scenarios. These geometric copies enable fast computational verification compared to analyzing the full complexity of actual hazard situations, improving verification productivity
3Reliability
If flight plans are verified in advance, then safety can be ensured, but uncertainties in measurements and predictions cannot be accommodated
Solution Approach 1:
The safety monitor applies different verification criteria to different regions of the flight path. By analyzing occupancy regions locally along the flight trajectory and comparing them segment-by-segment, the system can account for spatial and temporal variations in uncertainty while maintaining overall safety verification
Solution Approach 2:
The system extends verification from traditional 3D spatial analysis to 4D space-time occupancy regions. This additional temporal dimension allows the system to accommodate uncertainties in predictions and measurements by verifying safety across both space and time, not just at static points
Data Source
AI summary
A safety monitor arranged to: receive a flight plan comprising a plurality of waypoints for an aircraft to follow on a flight mission; receive hazard data corresponding to one or more hazards; generate an aircraft occupancy region comprising a four-dimensional volume corresponding to a range of possible spatial and temporal coordinates for the aircraft along the flight plan; generate a hazard occupancy region comprising a four-dimensional volume corresponding to a range of possible spatial and temporal coordinates associated with the one or more hazards; and determine whether there is an overlap between the four-dimensional volume of the aircraft occupancy region and the hazard occupancy region.