The data comes in every day – petabytes of it, streaming from instruments on satellites. Computers convert them into images, and analysts use those images to help make decisions in a wide range of areas including national security, weather prediction and agriculture. But those images alone only go so far.
“While a picture may be worth a thousand words, at the end of the day, a limited amount of intelligence can be pulled from a picture,” said Todd Kaiser, technical director, Constellation Management and Protection for Raytheon Intelligence & Space, a Raytheon Technologies business. “But, by fusing data from the picture with intelligence gathered from other sources, analysts can get a fuller picture, so to speak.”
That approach is called multiple intelligence, or multi-INT, sense-making. It includes fusing different types of data from a variety of sources to bring analysts greater clarity and enable better-informed decisions. RI&S is working with both government agencies and civilian firms to develop and deploy a new generation of custom-made multi-INT sense-making technologies that use cost-effective, high-performance computing integration, big-data storage and complex algorithms.
“Regardless of the source, if data isn’t quickly fused together and made readily available for analysts to get insights and make decisions, then it’s just noise,” said Kaiser. “Multi-INT sense-making technology helps provide actionable information, enabling our warfighters and intelligence analysts to make the right decisions faster than their adversaries.”
RI&S is developing machine-learning algorithms that can process and fuse petabytes of data in seconds to generate high-level interpretations, telling analysts broadly what the data means and making their jobs easier. From there, analysts can develop an informed hypothesis, providing information on how the data impacts a decision.
“The benefits of multi-INT sense-making go beyond national security,” said Tom Jones, technical director, Intelligence Production Solutions for RI&S. “For example, farmers receive reports providing geo-mapping fused with soil data from sensors. They then take that fused data to determine the health of their crops across their farm, and determine exactly where fertilizers or pesticides are most needed. This ensures the production of more crops, reducing costs and saving resources.”
While there are no entirely autonomous commercial farms today, there are several prototypes of automated farm equipment under development. When that equipment comes online, it will give farmers even more sources of data they can use to make better business decisions.
While more sensors mean more data, and while multi-INT sense-making helps turn those data into insights, even the most powerful technologies rely on humans to verify the findings and make the final decisions.
“The goal is not to completely eliminate humans from data-driven workflows, but to integrate computers deeper into the process, automating the fusion of data,” Jones said. “This helps analysts provide better intelligence based on their expertise, ensuring more informed decisions.”