Shovel Overhead View Collision Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Industrial machines like electric rope shovels face collisions with objects in their swing path, leading to potential damage to the machine and other objects, due to the lack of effective detection and mitigation systems.
Innovation Solution
A system comprising a processor that receives data from sensors to detect objects around the machine, identifies planes associated with predetermined objects like haul trucks, and superimposes these planes on an overhead view image, providing alerts and augmenting control to prevent collisions through audible, visual, and haptic feedback.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If sensors and detection systems are installed on the shovel to detect objects in the swing path, then collision detection capability is improved, but device complexity and cost increase
Solution Approach 1:
The sensor system is designed to detect multiple types of objects (haul trucks, personnel, equipment, environmental features) using the same sensor array and processing unit. The system can identify different object types by analyzing plane configurations and spatial relationships, allowing one detection system to serve multiple safety functions rather than requiring separate specialized sensors for each object type.
Solution Approach 2:
The system uses an intermediary processing layer that converts raw sensor data into meaningful object identification and collision risk assessment. The processor analyzes sensor readings to identify planes and their configurations, serving as a mediator between the physical sensor array and the control system, thereby simplifying the overall architecture while maintaining high detection capability.
2Measurement precision
If the system pre-identifies and stores plane configurations for all possible objects, then object detection accuracy is improved, but the amount of preprogramming and data storage requirements increase
Solution Approach 1:
The system segments the detection space into discrete planes rather than storing complete 3D models of objects. By identifying and tracking individual planes and their spatial relationships, the system achieves accurate object detection with minimal data storage requirements. Each plane is represented by simple geometric parameters rather than extensive point cloud data or mesh models.
Solution Approach 2:
The system uses lightweight plane representations instead of heavy pre-programmed object libraries. Rather than storing and matching against comprehensive 3D models of every possible object, the system creates simple plane geometries from sensor data that can be quickly processed and discarded, achieving high detection accuracy with minimal computational and storage resources.
3Reliability
If the system provides multiple alert methods (audible, visual, haptic feedback) and augment control, then collision mitigation effectiveness is improved, but device complexity and power consumption increase
Solution Approach 1:
The alert system uses periodic or pulsed feedback rather than continuous operation. Visual alerts flash or pulse, audible alerts use intermittent beeps or horns, and haptic feedback provides periodic vibrations. This periodic action pattern maintains high alert effectiveness while significantly reducing power consumption compared to continuous activation of all alert mechanisms.
Solution Approach 2:
The system dynamically selects and adjusts alert based on the specific collision risk scenario. The control system adapts the type, intensity, and combination of alerts (visual, audible, haptic) based on the detected object type, distance, and relative motion. This dynamic adaptation ensures effective mitigation while minimizing unnecessary power consumption from overly aggressive or redundant alerts.
Data Source
AI summary
Systems and methods for providing an overview-head view of an industrial machine, such as a shovel. One system includes at least one processor configured to receive data from at least one sensor installed on the shovel relating to the area around the shovel, identify a plurality of planes based on the data, determine if the plurality of planes are positioned in a predetermined configuration associated with a haul truck, and if the plurality of planes are positioned in the predetermined configuration, superimpose the plurality of planes on an overhead-view image of the shovel and the area.