This post is also available in: עברית (Hebrew)
A new, multiyear commitment to support the integration of modern “smart” technology into existing systems critical to the country’s defense was launched by Indiana University and the U.S. Navy. Through the cooperative research and development agreement, scientists at the IU School of Informatics and Computing and at the Naval Surface Warfare Center, Crane Division will work together to transform existing military sensor technology through machine learning and artificial intelligence, as reported by the university’s website.
The agreement is part of a larger effort at the IU School of Informatics and Computing to foster partnerships with NSWC Crane, a major economic driver in southern Indiana and one of the largest naval bases in the country.
“Artificial intelligence, machine learning and human-computer interaction are three areas of interest to the researchers at Crane, and also areas of great strength at our school,” said Martina Barnas, assistant dean for research and director of research collaborations at the IU School of Informatics and Computing. “We’re ideally positioned to assist their efforts in this important arena.”
“Machine learning refers to the development of algorithms that enable computers to learn and adapt to new situations based upon existing data,” said Sriraam Natarajan, one of the lead researchers. He said the collaboration between IU and NSWC Crane will apply these models to the massive amounts of data collected by naval technology — such as drones or radar arrays — in order to translate the information into a form that’s more useful to people.
Access to the Navy’s data is critical to the project because the more real data IU researchers can provide their machine learning system, the “smarter” it can grow. A typical naval monitoring system might use over 100 sensors to detect a wide range of phenomenon, such as infrared light and microwave emissions.
“With sensor technology growing so advanced and proliferating so quickly, human operators are unable to keep up with the information coming in..” “Increasingly, we need to augment human operators’ intelligence with artificial intelligence.”
To illustrate how machine learning strengthens sensor technology, Natarajan uses the example of two points of light traveling down a darkened road. Although a sensor can pick up the lights, it can’t necessarily determine the source of the light in the same way as a person, who can potentially interpret the light as a single vehicle or two motorcycles based upon whether the distance between the lights varies. A computer would need to use machine learning to make a similar assessment.
Other scenarios that could benefit from the technology include identifying objects on the ground based upon data collected by a drone in flight, identifying the type of object entering a roadway, or assessing the potential danger of an incoming projectile.
“The final decision always remains in the hands of the human operator,” said Robert Cruise, Chief Scientist at Crane. “Once the sensor data is understood, or a threat recognized, an artificial intelligence system turns that information over to the person in charge of the situation.”