Ten days of US military exercises in the California desert exposed shortcomings in several drones, communications systems and sensor-equipped helmets during a large-scale test of artificial intelligence and data integration technologies in combat conditions simulating a confrontation with the Chinese military.
9,700 troops take part in combat simulation
Around 9,700 troops took part in Project Convergence, a round-the-clock combat simulation against a US Army unit playing the role of the Chinese People’s Liberation Army, according to The Wall Street Journal. The exercises aimed to assess the ability of modern technologies to operate in a harsh desert environment under continuous field pressure. The tests included artificial intelligence applications, data-integration systems, aerial and ground drones, and sensor-equipped helmets.
However, several drones and communications systems failed to operate as expected, while soldiers operated some drones for limited periods without making full use of their autonomous capabilities or artificial-intelligence-based tools. During the exercises at Fort Irwin in California, a drone produced by Anduril Industries was forced to land after overheating.
Other drones lost contact with their operators when they moved behind mountains, highlighting the impact of terrain and high temperatures on the reliability of the systems in use. A drone produced by Thales UAV Solutions faced a lengthy delay before take-off, as its operators waited for its battery to finish charging, after which it needed an additional 10 minutes to take flight.
By comparison, Ukrainian soldiers can launch drones of the same size in less than a minute, according to the comparison included in the exercise results. Alex Miller, chief technology officer to the US Army chief of staff, said the exercises gave the Army an opportunity to test companies’ claims when presenting their products and to verify in practice how well these systems could meet the field needs of troops.
Augmented-reality helmet faults and costs
The problems extended to augmented-reality helmets, with tests revealing shortcomings in their performance. Three companies had received contracts worth more than 100 million dollars each to manufacture the helmets, while the US Army had not yet decided whether to continue using them.
General Michael McCurry, one of the senior commanders overseeing the exercises, said operating electronic devices in temperatures reaching around 43 degrees Celsius helped reveal technical problems that might not appear under standard testing conditions. The tests are part of US efforts to expand the military’s reliance on technology.
Calls to accelerate the adoption of artificial intelligence
Last year, US Defence Secretary Pete Hegseth called for transforming the Army into a force reliant on artificial intelligence, using drones and three-dimensional printers to produce weapons components rapidly, while increasing reliance on software and technologies developed by start-ups.
Hegseth stressed that the Army should not wait for technological perfection before bringing new systems into service, instead calling for the purchase of off-the-shelf products capable of meeting military needs to accelerate modernisation and test technologies in practical use. US investments in these fields have begun to appear in military operations.
During the Iran war, the United States used an unmanned boat produced by Saronic to help rescue the crew of a stranded helicopter. Artificial intelligence tools also helped accelerate intelligence analysis and target selection. Developments on the ground in Ukraine have reinforced the importance of drones and artificial intelligence in modern warfare.
A military-use gap compared with Ukraine and beyond?
Nevertheless, current and former national security officials believe the US Army remains far from making full use of autonomous weapons and artificial intelligence systems, compared with armies whose circumstances of war have driven them to accelerate the modernisation of their capabilities.
The war in Ukraine has again highlighted weaknesses in US military capabilities after Ukrainian forces demonstrated a high level of proficiency in using drones and modern technology. During military exercises held in Germany earlier this year, Ukrainian drone operators managed to take a US armoured brigade out of the simulated battle.
Despite the setbacks that emerged during the exercises, the US Army continues to develop command-and-control systems, foremost among them the next-generation command-and-control system, which uses artificial intelligence for planning and logistics. The system costs 4 billion dollars under this year’s budget.
The system is intended to accelerate the movement of military units and enable forces to deploy across wider areas of the battlefield, in line with the requirements of modern warfare. Starlink connectivity helped improve communication between soldiers and operations centres during the tests. Data-sharing technologies also reduced by two-thirds the time needed to plan missions and relay them to troops, according to military officials.
The next challenge is to improve the reliability of communications and data-integration systems so that artificial-intelligence-supported weapons can operate in remote areas and difficult field conditions.
