Working Machine Camera Collision Avoidance for Animate Object Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Off-highway vehicles or working machines, such as telescopic handlers, face challenges in detecting and avoiding collisions with workers or animals due to limited visibility during maneuvers, particularly during reversing, slow speed operations, loading/unloading, and implement coupling/uncoupling, leading to potential accidents.
Innovation Solution
A working machine equipped with a collision avoidance system featuring multiple cameras and a machine learning algorithm to detect animate objects, providing alerts and warnings to the operator through an array of indicators and a display, with zone-based risk assessment and priority display modes to enhance visibility and reduce collision risks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple cameras and machine learning algorithms are added to detect animate objects, then collision avoidance capability is improved, but device complexity increases
Solution Approach 1:
The system divides the monitoring area into multiple zones (first zone, second zone, third zone, fourth zone) corresponding to different camera fields of view. Each camera monitors a specific zone, and the control system processes each zone independently, allowing parallel processing and reducing overall system complexity while maintaining comprehensive monitoring coverage.
Solution Approach 2:
The control system acts as an intermediary that receives output signals from multiple cameras, processes them through machine learning algorithms to determine if animate objects are present, and generates appropriate output signals. This centralized intermediary coordinates the complex interactions between multiple sensors and actuators, managing system complexity while enabling intelligent collision avoidance.
2Area of stationary object
If cameras are positioned to cover all work areas, then detection coverage is improved, but the mounted positions become fixed and limit field of view
Solution Approach 1:
The system compensates for the fixed two-dimensional mounting positions of cameras by introducing a temporal dimension through machine learning algorithms. The algorithms process sequences of images over time, enabling the fixed cameras to detect and track moving animate objects effectively, thus achieving adaptive monitoring coverage without physical repositioning.
Solution Approach 2:
Each camera is designed to monitor a specific zone (first zone, second zone, third zone, or fourth zone) corresponding to different work areas. The control system universally processes output signals from all cameras using the same machine learning algorithms, allowing the system to adapt to different working conditions and animate object positions across all zones without requiring different detection methods for each area.
Data Source
AI summary
A working machine includes a body, a ground engaging propulsion structure supporting the body, a working arm, a control system, and a collision avoidance system. The collision avoidance system has a camera configured to monitor an area in its field of view and to generate an output signal to the control system in response to a detected animate object. The control system includes a processor configured to execute a machine learning algorithm trained to determine whether the detected object is an animate object to determine whether an animate object is within the field of view. The system provides an output to alert an operator of the working machine if it is determined that an animate object has been detected in the field of view of the at least one camera.