Authors : Mohan Doss Nadarajan; Sai Raghava; Sandeep Giri; Ranjitha.P; Bharani Kumar Depuru

Volume/Issue : Volume 8 - 2023, Issue 11 - November

Google Scholar : http://tinyurl.com/4pp4x9r5

Scribd : http://tinyurl.com/yc6ec2xc

DOI : https://doi.org/10.5281/zenodo.10391024

Abstract : Wooden pallets are widely used in the supply chain, yet they are susceptible to damage during storage and transportation. This susceptibility shortens the pallets' service life and results in significant costs due to product loss and pallet replacement. Automated pallet inspection can play a crucial role in identifying and preventing damaged pallets from entering the supply chain. Machine learning models, such as CNNs, SVMs, VGG16, VGG19, MobileNet, DenseNet, and ResNet51, have emerged as promising new approaches for automated pallet inspection. These models can be trained to automatically identify and classify pallet damage from images. This is achieved by training the model on a large dataset of labeled images of pallets with different types of damage, such as good, repair, and dismantle. Once trained, a machine learning model can swiftly and accurately classify new pallets. To do this, one simply feeds the model an image of a pallet and receives a prediction of the pallet's damage status. The predicted results are then stored in a database for future reference and analysis. The objective of this research is to develop an automated pallet inspection architecture with three key categories: 'good,' 'repair,' and 'dismantle.' The architecture will be based on a machine learning model, such as a CNN, SVM, or ResNet, trained on a substantial dataset of labeled images of pallets with various types of damage. Once trained, the model will be deployed in a real- time system for inspecting new pallets. The system will rapidly and accurately classify pallets and provide recommendations for categorizing them as 'good,' 'repair,' or 'dismantle.' The automated pallet inspection architecture holds the potential to enhance the efficiency and accuracy of pallet inspection, reduce reliance on manual inspection, and effectively identify and prevent damaged pallets from entering the supply chain. This, in turn, can lead to substantial cost savings and reductions in product loss.

Keywords : Automated Pallet Inspection, Artificial Intelligence, CNNs, SVMs, ResNets, Pallet Classification, Pallet Damage Classification, ResNet for Pallet Inspection, Pallet Dismantle Prediction, Real-Time Pallet Inspection, Wooden Pallet Quality Control.