Working Machine Vision Sensing for Moveable Element Position
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing working machines, such as backhoe loaders, face challenges in accurately determining the position of moveable elements due to the fragility and high cost of sensors, which are further compromised by dirty and dusty operating environments, leading to increased maintenance and operational costs.
Innovation Solution
A working machine equipped with a camera and a processor executing a machine learning algorithm, specifically a neural network, to determine the position of moveable elements from images, eliminating the need for sensors and improving reliability and reducing costs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sensors (linear encoders or rotary encoders) are used to measure the position of moveable elements, then measurement precision is improved, but device complexity and manufacturing cost increase
Solution Approach 1:
The patent replaces mechanical sensors (linear encoders and rotary encoders) with a vision-based system using a camera and machine learning algorithm. The camera captures images of the moveable element, and the machine learning algorithm processes these images to determine position, eliminating the need for complex mechanical sensing components while maintaining measurement capability
Solution Approach 2:
The patent creates a visual copy (image) of the moveable element and processes this copy to extract position information. Instead of directly measuring the physical element with sensors, the system captures an optical representation and uses computational methods to derive positional data from the image
2Measurement precision
If sensors are used to determine moveable element position, then measurement precision is improved, but reliability deteriorates due to fragility and susceptibility to damage
Solution Approach 1:
The patent replaces fragile mechanical sensors with a robust vision-based system. The camera and machine learning algorithm are less susceptible to physical damage and environmental factors, improving system reliability while maintaining the ability to accurately determine moveable element position
Solution Approach 2:
The patent employs a camera system that is generally more durable and less expensive than precision sensors. While cameras can be damaged, they are typically more robust to environmental conditions and can be replaced more easily than delicate encoder systems, effectively treating the sensing component as a more resilient alternative
3Measurement precision
If sensors are used in dirty and dusty environments, then measurement precision is maintained, but object-affected harmful factors increase due to dirt and dust obscuring sensor operation
Solution Approach 1:
The patent replaces contact-based mechanical sensors with a non-contact optical system. The camera captures images remotely, and the machine learning algorithm processes the visual data, eliminating the need for physical interaction between the sensing element and the measured object. This reduces susceptibility to dirt and dust accumulation that would obscure sensor operation
Solution Approach 2:
The patent introduces light as an intermediary between the camera and the moveable element. Instead of direct physical contact or close proximity sensing that would be blocked by dirt and dust, the system uses optical signals that can traverse the environment and carry information about the object's position without being significantly degraded by typical environmental contaminants
Data Source
Figure 1
Figure 2
Figure 3a
AI summary
A working machine (200) has a body and a moveable element. The moveable element is moveable relative to the body. A camera (242) captures an image of at least a portion of the moveable element. A processor (244) executes a machine learning algorithm trained to determine a position of a moveable element from an image of the moveable element.The machine learning algorithm receives the image from the camera (242) and determines the position of the moveable element in the image.