While NASA’s relationship with IBM is nothing new, a recent partnership with IBM’s Watson could help propel NASA research beyond its current human limitations. Watson, developed by IBM’s DeepQA project, is the computer system famously known for beating Jeopardy champion Ken Jennings in 2011.
One may presume that Watson is nothing more than a fancy search engine; however, Watson functions very differently than a search engine, in that it can more realistically parse natural language. Querying a search engine with too much information causes it to return several pages of potential results that may or may not be what a user needs; however, like human interaction, the more data you provide Watson, the more relevant and precise its response.
NASA isn’t the first organization to employ Watson for research efforts. This system is already spearheading cancer research projects within the healthcare industry. Just like healthcare, combing through thousands of aerospace research documents can also become tedious for humans. However, Watson acts as a research development advisor, analyzing and providing feedback on documentation written in several different languages, deconstructing complex mathematical equations, and conveying all of this data in a concise and easily digestible format for human consumption.
NASA conducted two pilot programs with Watson. The first examined its capacity to consume 130,000 documents on the topic of nanotube technology; and the second concentrated on Watson’s ability to clearly communicate its findings to NASA engineers and researchers. The results of these pilot programs were validated by Subject Matter Experts (SMEs) and it was concluded that Watson made connections, which would have likely been overlooked by human researchers. In addition, these pilots assisted in the development of a more straightforward User Interface (UI) for the system.
In the future, NASA anticipates Watson can help them answer questions and solve spaceflight science problems in real-time—aiding the organization in making vital decisions mid-flight that could both save millions of dollars in equipment and human lives.