Army researchers seek more data for soldiers with two projects
WASHINGTON – The U.S. Army Research Lab made inroads this summer on two neural network projects that could help commanders make battlefield decisions and provide information about soldiers’ health through fiber of their uniform.
The advancements come as the U.S. military prepares for a data-driven battle, in which masses of data are transmitted across the battlespace, processed, and used in a commander’s decision-making. Neural networks are a combination of algorithms that work together to recognize patterns of data through a process similar to that of the human brain.
The first project aims to provide a tool for battle commanders that quantifies uncertainty in data analysis using neural networks. Researchers associated with the Army Research Lab have created a new framework for processing neural networks that would use artificial intelligence to provide confidence ratings. The tool could improve the ethical use of AI and provide commanders with a new level of confidence in their decision-making.
“The idea is to provide some sort of measurement of uncertainty of the AI system so that human counterparts are able to recognize and realize that the AI will not be perfect, but we can sort of understand to how certain it is, then use that information to make more informed decisions, ”said Maggie Wigness, Army researcher and deputy director of the Internet of Battlefield Things Collaborative Research Alliance (IoBT CRA ).
Several factors influence the uncertainty about battlefield data and AI outputs, Wigness said. The data could be manipulated, it could have random “noise” that disrupts algorithms, or uncertainty could come from a degraded sensor that an opponent has attacked. To quantify uncertainty in the future, researchers classified sources of uncertainty in military networks, examined frameworks to represent the unknown, and created solutions to manage it.
As the military strives to enable multi-domain operations, in which the service faces adversaries in all five combat domains, data from the five domains will need to be merged and integrated. The work on the neural network fits into the work of the IoBT alliance which is “focused on processing data, establishing connections and communicating through a large network of sensors, these sensors could be in some way. so anywhere in those multiple areas that then need to be merged, ”Wigness mentioned.
Meanwhile, Army researchers at the Institute for Soldier Nanotechnologies at the Massachusetts Institute of Technology developed a programmable fiber that could be sewn into soldiers’ uniforms in the future and provide biometric data that would help monitor the health of soldiers. This fiber, depending on the service, can detect, store, analyze and infer activity. Digital fiber technology is in the proof of concept phase, demonstrating that fiber has memory capabilities. The researchers loaded a film onto the fabric to show that it could store massive amounts of information.
In the future, researchers want the fiber to be able to store algorithms capable of analyzing the data it collects. According to an ARL article on the breakthrough, the fiber has a neural network of 1,650 connections and successfully collected 270 minutes of surface body temperature data. Using artificial intelligence in the lab, the fiber identified the type of activity the soldier was undertaking with 96% accuracy.
Ultimately, researchers want the fiber to be able to generate and store energy to perform artificial intelligence analyzes or communicate with sensors and communication systems.
“You can imagine that if a soldier doesn’t respond, we might be able to communicate with these laptops… where the soldier’s uniform communicates with central command or his home base on the battlefield,” James said. Burgess, Institute for Soldat Responsible for the Nanotechnologies program for the research laboratory. “So if they don’t respond, we can know… if they’re bleeding, if they’re breathing, what their pulse is, so we would have some diagnostic information before we even reach the soldier with the rescue team.” “
The capability is still “at a very early stage,” Burgess said, and the team aims to have an impact on the battlefield by 2050. The capability will also be an important component in providing soldiers with information about their position. physical state.
“With the adrenaline going up, it’s really important for soldiers to know what’s happening to them, their physiological state, and to have horns, basically, going off. [if] they overheat or become dehydrated if they’ve been exposed to something that begins to impair their cognitive or physical abilities, ”said Burgess.
Andrew Eversden covers all defense technologies for C4ISRNET. He previously reported on Federal IT and Cyber Security for Federal Times and Fifth Domain, and worked as a Congressional Reporting Officer for the Texas Tribune. He was also an intern in Washington for the Durango Herald. Andrew graduated from the American University.