According to the US Stars and Stripes website on April 30, Army officials recently said that a new artificial intelligence (AI) project may change the way the U.S. military manages its munitions demand. It achieves performance that is better than the current system by analyzing and processing a large amount of data.

For the United States, which lacks military production capacity, the practice of using AI to improve the efficiency of ammunition can be regarded as a tactical and active cover-up strategy.

For the United States, which lacks military production capacity, the practice of using AI to improve the efficiency of ammunition can be regarded as a tactical and active cover-up strategy.

The United Ammunition Command said in a statement that the program, known as the "quarterly resupply model," aims to revolutionize the logistics of the military's vast ammunition management system.

In a recent experiment, a machine learning model generated 27,300 forecasts, with a forecast accuracy of 74% for ammunition demand, while the current system's accuracy is only 25%.

Ryan Senkebier, acting director of the ammunition logistics distribution department of the command, said in a statement that the discovery “represents a significant advance in ammunition logistics.”

The Army said the milestone was achieved in September last year, after a year-long development process.

Joint Ammunition Command said the model could consider variables in ammunition consumption, training plans and installation requirements.

For many years, the military has relied on the "Gross Ammunition Management Information System" to handle the distribution of ammunition in various units.

This work is managed by the Army, but most of the U.S. troops rely on this. According to the Marine Corps' 2021 project summary, the current system calculates combat load requirements, validates and sends electronic requests, collects consumption data, and makes predictions.

It is a costly project for the military to ensure that the U.S. military and its troops around the world have ammunition. The Army has not stated whether AI tools will eventually replace the current system. But they say new machine learning methods can analyze historical data to better predict future ammunition demand.

Senkbier said the purpose of the new model is to reduce shipping frequency, optimize inventory levels, and improve prediction accuracy.

The Army said the model is operating at relevant domestic locations and its performance remains to be observed. The Army added that this data-driven approach “can be continuously improved.”