Pallet Placement Simulation for Stable Mixed-Item Stacking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for palletizing and depalletizing heterogeneous items are inefficient due to the manual selection and stacking of items, which can result in unstable pallets and increased resource intensity.
Innovation Solution
A robotic system that uses a simulation service to determine optimal placement locations for items on a pallet by iterating over different placement models and state estimation models, ensuring stability and efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual selection and stacking of items is used, then human judgment and intuition can be applied to place heavier items on the bottom, but the process is resource-intensive and results in unstable pallets
Solution Approach 1:
The patent replaces manual human operations with an automated robotic system that uses computer vision and machine learning algorithms to select and place items. The robotic system substitutes human judgment with automated image processing and stability prediction models, eliminating manual labor while maintaining or improving pallet stability through computational analysis of item attributes and placement configurations.
Solution Approach 2:
The system performs preliminary simulation of multiple placement scenarios before actual palletization. By predicting stability outcomes and evaluating different stacking configurations in advance using machine learning models, the system determines optimal placement sequences beforehand, ensuring stable pallets are constructed efficiently without trial-and-error manual adjustments.
2Ease of operation
If items are selected from bins in an ordered list, then the process is simplified, but the palletized set becomes unstable
Solution Approach 1:
The robotic system continuously captures images of items in bins and the developing pallet stack, processes this visual feedback through machine learning models, and adjusts placement decisions in real-time. The system uses feedback from predicted stability metrics to modify the ordered list approach, dynamically reordering item selection to maintain stability while preserving automated simplicity.
Solution Approach 2:
The system transforms the static ordered list into a dynamic selection process that adapts based on current pallet configuration and predicted stability outcomes. The item selection order is not fixed but dynamically adjusted by the machine learning model based on real-time assessment of the stacking situation, combining automated simplicity with adaptive stability management.
3Extent of automation
If robotics is used to palletize items, then automation and efficiency are improved, but the variety of items and variations in order, number, and mix make the process more challenging
Solution Approach 1:
The machine learning model is trained to handle variations in item parameters such as size, shape, weight, and packaging type. By adjusting the model's parameters and training data to account for diverse item attributes, the system maintains high automation levels while adapting to different item varieties, orders, and mixes through computational rather than mechanical adaptation.
Data Source
AI summary
A robotic system is disclosed. The system includes a memory that stores for each of a plurality of items a set of attribute values. The system includes a processor(s) that uses the attribute values to simulate the placement of items, including by determining, iteratively, for each next item a placement location at which to place the item on a simulated stack of items on the pallet, using the attribute values and a geometric model of where items have been simulated to have been placed to estimate a state of the stack after each of a subset of simulated placements, and using the estimated state to inform a next placement decision. The steps of determining for each next item a placement location and estimating the state of the stack until all of at least a subset of the plurality of items have been simulated as having been placed on the stack.