Over the past months, since the last update on the project, progress has been made on the development of the artificial intelligence algorithms required for the optimization of the boxes that are placed on pallets. Heuristics have been obtained to adjust the hyperparameters of the neural networks. Then, populations of individuals have been simulated with neurogenic algorithms, improved through different generations.
As can be seen in the image below, populations of 500 individuals have been simulated, which have been improved with iterations.
These experiments have resulted in an initial version of the neurogenic algorithm that is able to calculate the best position for the placing of boxes of products on the shipping pallet. This information is provided to the operator via an application interface on their mobile device.
The calculation times of the algorithm are expected to reduce in future versions. Moreover, an assessment will be made of the impact of the application on the operators' work. The image shows a simulation of how the application indicates the position in which the product should be placed inside the shipping pallet so that it is not damaged on the route and the number of pallets to be used in these shipments is reduced.