Cell-Based Object Detection for Mobile Collision Avoidance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems fail to effectively detect moving objects and prevent collisions in dynamic environments, as they often rely on incomplete or inaccurate data from individual sensors, leading to inefficient collision avoidance strategies.
Innovation Solution
A method and system where multiple handsets perform measurements and transmit data to a higher-level computer, which evaluates the data to create a comprehensive map, allowing for centralized determination of moving objects and adjusting trajectories to avoid collisions, utilizing a laser scanner and odometric position determination for efficient object tracking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple handsets perform measurements and transmit data to a higher-level computer for centralized evaluation, then the accuracy of moving object detection is improved, but the system complexity increases
Solution Approach 1:
The system divides the detection task among multiple handheld devices (segmentation of detection function), where each device independently performs measurements and transmits data to a higher-level computer for centralized evaluation. This segmentation improves detection accuracy through multiple data sources while distributing system complexity across independent units.
Solution Approach 2:
A higher-level computer acts as an intermediary that receives measurement data from multiple handsets, performs centralized evaluation to distinguish moving from stationary objects, and coordinates collision avoidance actions. This intermediary consolidates complexity in a single coordination point while enabling accurate multi-device data integration.
2Reliability
If the system initially classifies all detected objects as moving objects, then collision avoidance is improved, but the precision of object classification deteriorates
Solution Approach 1:
The system performs preliminary classification by initially treating all detected objects as moving objects, ensuring conservative collision avoidance from the outset. This preliminary action prioritizes safety before refined classification is achieved through multiple measurements and higher-level computer evaluation.
Solution Approach 2:
The system uses feedback from multiple measurements and higher-level computer evaluation to progressively refine object classification. As more measurement data becomes available, the system transitions from conservative preliminary classification (all objects as moving) to more precise classification based on accumulated evidence.
3Loss of information
If a central computer compiles and analyzes data from all measurements, then the completeness of object detection is improved, but the data transmission requirements increase
Solution Approach 1:
The higher-level computer merges measurement data from all handheld devices into a unified detection result, achieving complete object detection information. This consolidation combines distributed measurements into comprehensive detection coverage while managing data transmission through centralized processing.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables accurate identification of moving and immovable objects, allowing for proactive collision avoidance by predicting object trajectories and adjusting target paths, thereby enhancing safety and operational efficiency in dynamic environments.
Implementation Method 1
The emitted laser beam is at least partially backscattered or reflected by the respective object (2, 3, 4), so that the direction and distance of the object (2, 3, 4) relative to the laser scanner can be determined.
Implementation Method 2
The emitted laser beam is at least partially backscattered or reflected by the respective object (2, 3, 4), so that the direction and distance of the object (2, 3, 4) relative to the laser scanner can be determined.
Implementation Method 3
The positioning system works, for example, odometrically
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The invention relates to a method for detecting moving objects in a system and/or for preventing collisions in a system and to a system for carrying out a method of this kind, wherein: the travel surface for mobile parts of the system is or has been divided into cells; each cell is uniquely assigned to one surface region of the travel surface; a first and a second counter is allocated to each cell; during each measurement performed by a particular mobile part using sensors - the cells on which an object was detected are determined and the count of the first counter for these cells is incremented, - and the cells on which no object was detected are determined, and the count of the second counter for these cells is incremented; for each cell the quotient of the count of the first counter and the count of the second counter is calculated; and depending on the value of the quotient for the particular cell - the presence of a non-moving object is assigned.