Automated Package Registration Using MVRs for Unknown Object Handling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Robotic systems face challenges in accurately registering and handling unknown or unrecognized objects, leading to potential mishandling and increased risk of injury to human workers during de-palletization, as they often rely on pre-registered data and may fail to correctly identify physical characteristics like dimensions and weight.
Innovation Solution
The robotic system employs image processing and sensor data to generate Minimum Viable Regions (MVRs) for unrecognized objects, using 2D and 3D imaging to identify exposed edges and corners, and autonomously registers these objects by generating verified MVRs, allowing for accurate manipulation and registration without initial data or human input.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the robotic system relies on pre-registered data for object handling, then the system operation is simplified and faster, but the system cannot handle unrecognized objects and may mishandle them leading to safety risks
Solution Approach 1:
The system performs preliminary actions by capturing images of objects on the pallet before handling them. The imaging device captures images of the pallet and its objects, and the system processes these images to identify physical characteristics such as dimensions, weight, and center of mass. This preliminary imaging and identification allows the robotic system to prepare handling parameters in advance, enabling both speed and adaptability.
Solution Approach 2:
The system uses the existing image data from the pallet capture process to automatically register new objects without requiring separate registration steps. When an unrecognized object is detected, the system extracts its physical characteristics from the already-captured images and uses this information to determine appropriate handling parameters, allowing the system to self-register and handle new objects autonomously.
2Measurement precision
If the system processes every object through manual registration, then accurate physical characteristics are obtained, but the process becomes resource-intensive and time-consuming
Solution Approach 1:
The system applies partial processing by selectively handling different types of objects differently. Recognized objects are handled using pre-stored parameters without re-processing, while only unrecognized objects undergo full image processing and registration. This partial action approach maintains measurement precision for new objects while avoiding redundant processing of known objects, significantly reducing overall registration time.
Solution Approach 2:
The imaging system serves multiple functions: it captures images for both recognized and unrecognized objects, extracts physical characteristics for registration, and provides data for handling parameter determination. This multi-functionality allows the system to obtain accurate measurements for new objects through a single imaging process rather than requiring separate registration steps, reducing time loss while maintaining precision.
3Reliability
If the robotic arm uses pre-stored handling parameters, then handling is consistent and safe, but the system cannot adapt to new object types with different physical characteristics
Solution Approach 1:
The system implements feedback by continuously comparing captured object images against the database of recognized objects. When an object is identified as unrecognized, the system processes its images to extract physical characteristics, determines appropriate handling parameters, and stores this information for future use. This feedback loop ensures that the system maintains reliability by using data-driven handling parameters while simultaneously improving adaptability by learning from new object types.
Solution Approach 2:
The system performs preliminary identification and parameter determination for unrecognized objects before actual handling occurs. By capturing images and extracting physical characteristics in advance, the system prepares appropriate handling parameters for new object types before they need to be manipulated, ensuring both safety through proper parameter selection and adaptability through automatic recognition and registration.
Data Source
AI summary
The present disclosure relates to detecting and registering unrecognized or unregistered objects. A minimum viable range (MVR) may be derived based on inspecting image data that represents objects. The MVR may be determined to be a certain MVR or an uncertain MVR according to one or more features represented in the image data. The MVR may be used to register corresponding objects according to the certain or uncertain determination.