Autonomous Passenger Boarding Bridge Navigation via Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The complex environment of aircraft operating areas and non-standard designs of passenger boarding bridges pose challenges for automated navigation and collision avoidance of equipment devices such as passenger boarding bridges.
Innovation Solution
A computer-implemented method using an autonomous control computing system that receives images from digital cameras, employs machine learning models to detect self and intruder objects, predicts future locations of self objects, and alters navigation paths to prevent collisions with intruder objects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If automated navigation systems are implemented for passenger boarding bridges, then operational efficiency is improved, but the complexity of the control system increases
Solution Approach 1:
The patent replaces manual mechanical navigation control with an automated computer vision-based navigation system. The system uses cameras to capture images of the aircraft operating area, processes these images through machine learning models to detect objects and navigate the boarding bridge autonomously, thereby improving operational efficiency while managing system complexity through software-based solutions.
Solution Approach 2:
The navigation system enables the passenger boarding bridge to navigate and position itself autonomously without human intervention. The system independently detects its own position, identifies obstacles and targets, plans navigation paths, and executes movement commands, making the system self-sufficient and improving operational efficiency.
2Reliability
If machine learning models are used to detect objects in complex environments, then collision avoidance capability is improved, but computational requirements and system complexity increase
Solution Approach 1:
The patent replaces traditional mechanical sensors and simple detection algorithms with machine learning models that process camera images to detect objects, predict their movements, and assess collision risks. This substitution enables more accurate and reliable collision avoidance in complex environments by leveraging pattern recognition and predictive analytics from visual data.
Solution Approach 2:
The system continuously captures images of the aircraft operating area, processes these images through machine learning models to detect objects and their movements, predicts future positions, and adjusts the navigation path in real-time based on these feedback loops. This continuous feedback mechanism improves collision avoidance capability by adapting to dynamic environmental changes.
3Adaptability or versatility
If autonomous navigation is implemented without standardization, then adaptability to different aircraft types is improved, but the difficulty of detecting and measuring objects increases
Solution Approach 1:
The patent implements a universal machine learning model that can detect and recognize various types of objects in the aircraft operating area, including different aircraft models, ground vehicles, and equipment. The model is trained on diverse data to generalize across different object types and environments, enabling the system to adapt to various aircraft types and operational scenarios while maintaining consistent object detection capabilities.
Solution Approach 2:
The system dynamically adjusts detection parameters and navigation constraints based on the detected aircraft type and operational context. By changing parameters such as detection sensitivity, path planning constraints, and safety margins according to the specific aircraft configuration and environment, the system maintains high adaptability while managing detection difficulty through context-aware parameter optimization.
Data Source
AI summary
A system for automatically determining a distance to an object in a two-dimensional image of an aircraft operating area. The system includes one or more processors configured to acquire at least one image captured by at least one camera of the aircraft operating area, detect the object in the at least one image where an identifier of the object and a component of the object are visible in the at least one image, acquire a dimension of the component of the object based on the identifier, determine a size of the component in the at least one image, and determine the distance between the component of the object and the at least one camera based on the dimension of the component and the size of the component in the at least one image.