The National Aeronautics and Space Administration’s (NASA) Jet Propulsion Laboratory (JPL) isn’t just looking into rockets these days.
Tom Soderstrom, IT chief technology and innovation officer at JPL, said Tuesday at Amazon Web Services’ inaugural Earth and Space Day that the NASA research center uses artificial intelligence (AI), digital assistants, chatbots, and Internet of Things (IoT) integration to enhance operations and cybersecurity capabilities, which will eventually fuel even bigger breakthroughs.
Soderstrom said all organizations can take advantage of small but practical projects to improve security and user experience and contribute to the next great wave of technology.
JPL–an 80-year-old start-up–is NASA’s Federally-funded research and development center, boasting some 6,000 employees. Soderstrom calls himself a “futurist” who looks into “technology waves”–where three, five, and ten-year periods of advancement culminate. He said these waves are all leading to the next “technology tsunami,” which is “built-in intelligence, everywhere.”
Soderstrom said much of the advanced research in AI–the things that will eventually answer the era’s big questions–will first be driven by intuitive user experiences.
“How do you interact with computing? It’s about being able to type to it, click through it, touch it on the big screen, or not touch it, use gesturing. It’s about talking to it, hearing it, and then seeing it on a big display or through your augmented reality glasses, sensing it through thousands of IoT sensors, and soon think it, all at the same time,” Soderstrom said. “That makes for a great user interface. How do you make the experience powerful? You apply AI.”
He said that AI-powered intelligent assistants will be crucial to productivity gains in the future.
“If you look at the statistics, humans are about 80 percent effective, machines are about 80 percent effective. When you use them together, they’re about 95 percent effective.”
He gave the example of a simple chatbot–a room assistant–that JPL designed “overnight” and is now used by more than 1,000 people a week to automatically find conference rooms that suit their needs. Users can ask the bot about specific connectivity features in the room, give it size considerations, and the system provides all the available options.
JPL designs its assistants on open APIs, allowing users to interface with the assistants in any way they prefer, whether that be via multiple text formats or with their voices through Amazon Alexa integration.
JPL integrated conference rooms with Alexa, IoT sensors, and computers running various cybersecurity dashboards. The Amazon assistant now understands queries that ask it to turn on screens, expand those dashboards, and display data about security incidents.
Soderstrom showed a video demonstration that illustrates just how seamless and intuitive the process is: JPL employees are making requests to digital assistants the same way they would a normal person.
JPL has also designed a cybersecurity chatbot that can easily connect IT staff to relevant information like security operations center incidents, check IP blocks, and route users to service requests. Soderstrom said JPL is partnering with industry players to work on cell phone authentication for government credentials that operate in lieu of official badges.
Soderstrom also said that JPL developed another one-day prototype: facial recognition software crafted one day before JPL’s major IT expo. The AI software recognized participants and automated access to restricted areas at the expo. JPL also designed a system that measured traffic flow in and out of the conference floor, using cameras and software installed on Raspberry Pi single-board computers. He said this type of instant analytics, powered by AI, can and will be used in space missions.
But Soderstrom said the real-world application of these intelligent assistants isn’t rocket science and pertains to every organization. “It’s really about the combinations, trying things for your use case,” he said. After asking audience members which of three examples–IoT, AI, and voice as a platform–seem best for their organizations, he said he believes it’s all three.
“It’s a completely new world, and it’s all driven by cloud computing, and for us, it’s going to help us answer those big questions in my lifetime, so protect me so I live long enough to see it,” Soderstrom joked, providing one final charge to the audience.
“What can you do in your companies to move us forward through experimentation? Lightweight, easy, fast, and it’s an experiment. The only thing that can fail in an experiment? That you don’t learn anything,” he said.
