Train Length Determination Using Machine Vision Cameras
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Positive Train Control (PTC) systems rely on human input for validating train consist information, leading to potential errors in train length, which can result in unsafe operations, such as overrunning targets or fouling tracks, due to inaccuracies in brake algorithms and speed restrictions.
Innovation Solution
A method using machine vision techniques with cameras mounted on trains to determine train length by comparing images of track features with a database, ensuring accurate train length calculation and clearance confirmation, thereby reducing human error and enhancing safety.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If crew input is used to validate train consist information, then operational flexibility is maintained, but measurement precision of train length deteriorates due to human error
Solution Approach 1:
The patent replaces the manual mechanical process of crew measurement and validation with an automated optical system. Cameras mounted on the train capture images of trackside objects, and machine vision algorithms automatically calculate train length by identifying the positions of the locomotive and end-of-train device relative to known track features, eliminating human error while maintaining operational flexibility
Solution Approach 2:
The system enables the train to self-measure its own length using onboard cameras and processing equipment. The train automatically captures images, processes them through machine vision algorithms, and validates consist information without requiring external crew input, thereby improving measurement precision while maintaining ease of operation
2Device complexity
If manual validation of train consist information is used, then system complexity is reduced, but reliability deteriorates due to potential human mistakes
Solution Approach 1:
The patent substitutes manual validation processes with an automated machine vision system that uses cameras, image processing algorithms, and database comparisons to reliably determine train length and validate consist information, significantly improving safety assurance through objective, error-free automated measurement
Solution Approach 2:
The system continuously captures images from onboard cameras, compares detected trackside objects against a database of known features, and uses this feedback to automatically validate train consist information. This closed-loop feedback mechanism ensures high reliability by constantly verifying train length and position data against expected values
3Ease of operation
If train length is inaccurately determined, then ease of operation is maintained, but harmful factors increase due to unsafe operations like overrunning targets or fouling tracks
Solution Approach 1:
The system performs preliminary validation of train length and consist information before the train proceeds with operations. By pre-verifying accurate length measurements and clearing track obstructions using machine vision analysis, the system prevents unsafe operations such as overrunning signals or fouling tracks before they can occur
Solution Approach 2:
The patent replaces subjective human judgment with objective machine vision-based measurement to determine train length and detect track conditions, eliminating the harmful factor of human error that could lead to unsafe operations while maintaining ease of operation through automated decision support
Data Source
AI summary
In a method of determining a length of a train, a first camera at the head of the train acquires a first image of a first object on or proximate a path of the train and a second camera at the end of the train acquires a second image of a second object on or proximate the path of the train. A controller determines a length of the train based on a first geographical location associated with the first object in the first image and a second geographical location associated with second object in the second image. The first and second geographical locations of the first and second objects can be determined from corresponding prerecorded images of the objects that are geotagged with the geographical locations of the objects.