Virtual Perimeter Sensing for False-Positive Control in Work Areas
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing object detection systems in autonomous vehicles fail to distinguish between false positives and do not address the need to optimize the detection of autonomous systems, particularly in agricultural and forestry applications, where standing crops are challenging.
Innovation Solution
A distributed system and method for generating a virtual perimeter using smart sensing stations to optimize perception sensing, particularly in agricultural and forestry applications, where standing crops are challenging.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If long-range object detection is implemented continuously throughout the work area, then detection coverage is improved, but resource consumption and false positives increase
Solution Approach 1:
The work area is divided into multiple zones with different detection requirements. Long-range detection is only activated in specific zones where it is truly needed, while other zones use shorter-range detection or remain undetected. This segmentation allows the system to maintain adequate detection coverage while significantly reducing resource consumption by avoiding continuous full-area long-range scanning.
2Measurement precision
If object detection sensitivity is increased to detect all possible objects, then detection precision is improved, but false positives increase causing unnecessary work stoppages
Solution Approach 1:
Different detection sensitivity levels are applied to different spatial zones and object types within the work area. High sensitivity is used only in critical zones or for critical object types, while lower sensitivity is used elsewhere. This local quality approach maintains detection precision where needed while reducing false positives that would trigger unnecessary work stoppages.
3Extent of automation
If autonomous operation is implemented without virtual perimeters, then automation level is improved, but safety risks increase due to undetected objects
Solution Approach 1:
Virtual perimeters are established in advance to define safe work areas and boundaries before autonomous operation begins. These pre-defined perimeters allow the autonomous system to operate with clear safety guidelines, knowing which areas are authorized for operation and which require heightened caution or are prohibited. This preliminary action maintains high automation while reducing safety risks through pre-planned spatial constraints.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A system and method are provided for distributed perception sensing via an object detection station arranged such that an associated one or more object detection sensors have respective fields of perception including at least a portion of a virtually defined perimeter extending at least partially about a working area. The object detection station detects a motion incident corresponding to a violation of the virtually defined perimeter, wherein a detected source of the motion incident is classified with respect to one or more defined source types, and one or more signals at least corresponding to the motion incident are wirelessly communicated to one or more working machines operating within the working area. At least one of the working machines may then be selectively transitioned from a normal operating mode to a hierarchical intervention mode with respect to the detected source and/or a relative position thereof within the working area.