Selective Loading Automation for Work Vehicles Under Sensor Uncertainty
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Novice operators face challenges in safely approaching and loading materials into truck or hopper areas due to difficulties in accurately measuring and controlling attachment and vehicle movements, often resulting in potential collisions, especially with the presence of visual defects like reflections and fog affecting stereo camera systems.
Innovation Solution
A system and method that utilizes a detection system with vision sensors to identify loading areas and automate the loading process by receiving user inputs through a user interface, allowing for selective automation of work vehicle movements and attachment control, even in conditions where machine learning algorithms may not be reliable, ensuring safe and accurate loading operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If stereo camera systems are used to detect loading areas, then automated loading control is enabled, but measurement precision deteriorates due to visual defects like reflections and fog
Solution Approach 1:
The system introduces an intermediary human operator who verifies and corrects the stereo camera's detection of loading area parameters. When visual defects cause measurement uncertainty, the operator acts as a mediator to provide accurate distance and edge information, ensuring precise automated control despite degraded sensor performance.
Solution Approach 2:
The system implements feedback by allowing operators to review and correct detection results from the stereo camera system. This feedback loop compensates for measurement errors caused by reflections and fog, maintaining high precision in automated loading control by continuously refining the detected loading area parameters.
2Ease of operation
If full automation is implemented for loading operations, then operator workload is reduced, but system complexity increases requiring sophisticated machine learning algorithms
Solution Approach 1:
The system applies partial automation rather than full automation, automating only the loading control functions where stereo camera detection is reliable. For cases involving visual defects, the system selectively engages human operators, providing the necessary level of automation to reduce workload without requiring complex machine learning algorithms for all operating conditions.
3Reliability
If selective automation is implemented allowing operator intervention, then measurement reliability improves in challenging conditions, but operation time increases due to manual confirmation steps
Solution Approach 1:
The system applies partial automation by requiring operator intervention only when necessary - specifically when visual defects like reflections or fog degrade stereo camera performance. In clear conditions, the system operates fully automatically without operator confirmation, thus maintaining high reliability when needed while minimizing time loss by avoiding unnecessary manual steps during normal operations.
Data Source
AI summary
A system and method of selective input confirmation for automated loading by a work vehicle comprising a main frame and a work attachment movable with respect to the main frame for loading/unloading material in a loading area external to the work vehicle during a loading process having loading stages. Location inputs are detected for the loading area respective to the main frame and/or work attachment. First user inputs correspond to selected automation for respective loading stages, for which detection routines are executed with respect to parameters of the loading area based on the detected location inputs. If second user inputs are determined to be required with respect to certain parameters of the loading area, the second user inputs are received and movement of the main frame and/or work attachment are controlled for automating the corresponding loading stages based at least in part thereon.