Tom Soderstrom, the chief technology officer for the NASA Jet Propulsion Laboratory, has a theory about incorporating new technology: You can't know whether it will be useful until you actually try it first.
That's why Soderstrom, who is a featured speaker at the CITE Conference, is always prototyping new technology at the organization, which designs unmanned spacecraft like the Curiosity Mars sover for NASA. The process has allowed NASA JPL to get way ahead on big changes like cloud computing.
The prototyping process runs on a three-year cycle -- which Soderstrom calls an "IT decade" because of the rapid pace of change in information technology -- and the groups doing the testing are largely self-selecting. If users are excited about a technology and actually make use of it, that's a good sign.
"One thing we look at for quick prototypes is not return on investment, because they're very rapid and very small, but return on attention. If we get end users excited about a prototype using emerging technology, and it is successful, it's a slam dunk for doing an investment."
The hands-on process not only helps NASA JPL decide where to invest, it often uncovers key points about where new technologies can fit into its business -- the kind of information that technical assessments might miss.
"When the iPhone came out, we explored, we experimented, we prototyped it, and we realized it wasn't about the phone. It was about the apps. Now people don't think about it, but it's a drastically different business model," Soderstrom told CITEworld.
"In the old days, we'd go to Egghead and for a couple hundred dollars buy a big book with some CDs. We lived with that model for years. Now the half life of an app is zero to a month or two."
The iPhone also changed how people interacted with one another.
"When the iPhone came out, you'd have the CEO of the company willingly and happily spend time with interns because the intern and CEO would swap phones and say 'Oh, what apps do you use? Oh what's Yelp?' I've never seen anything like that happen before. Maybe when the first PCs came out, but the PC wasn't used by CEOs in those days. The iPhone democratizes and speeds up innovation. It's intriguing -- the same thing happened with the iPad."
This prototyping process also helped Soderstrom uncover some duds, like the first iteration of 3D printing -- the devices were too expensive at $10,000 apiece -- and Simple Object Access Protocol (SOAP), a way for applications to use XML to exchange information over the web. "That was too early. While we were talking about SOA [service-oriented architectures] and SOAP, people were moving to REST. Developers learned it, did it, and delivered it, while the SOA people were still talking. Whatever developers will use, which is whatever is simplest, is what will take off."
So what is Soderstrom seeing NASA JPL users get excited about in the current three-year test cycle?
- Data visualization. "Last summer we did a 'startup' on big data and big data analytics. We wrote our own visualization software. The visualization software was good, but somebody has to maintain it," said Soderstrom. "One of our students ended up using Tableau to do the visualization part, and everybody migrated over to that as opposed to what we'd custom written. We also use Splunk and some other things." He continues, "The big deal for us is to be able to give end users self-service analytics. Self service analytics means they can interact with their own data using something super easy to use, visually exciting, and works on mobile devices."
- The Internet of things. Soderstrom uses the concept of "dark data" to describe "data we already had but didn't care about, like email data, or data coming from outside that we could never take advantage of." The Internet of Things is an example of the latter. "We can very rapidly gain insight from data that we previously didn't know existed."
- 3D printing. "All of a sudden it's back. It flopped because $10,000 for a cheap 3D printer is not cheap enough. But $1,000 is cheap enough," said Soderstrom. "In one year, we've gone from it being a petting zoo experiment to actually being used in missions now. The way it's being used in missions is for brainstorming and design and rapid prototyping. You can see it, touch it, discuss it with scientists and engineers, and modify design. The cycle time has sped up tremendously." He acknowledges that 3D printing is not quite ready for the mass consumer market, but neither was the first generation of dot-matrix printers -- only when inkjet printers became common did regular printing really take off.
- Robots. "SMAC [social-mobile-analytics-cloud] are all key disruptors in their own right, but especially when it comes to consumer robotics. We're quite bullish on that."
- New more natural UIs. "Anything from Siri to in our clean room we're able to use touch on these big screens so we can redline drawings without printing in a clean room." As with 3D printing, dramatic price drops are going to drive touch into bigger and bigger screens, Soderstrom thinks.
In addition to these five technological areas, Soderstrom emphasized the importance of startups, which he called the "new gold rush. We participate in hackathons, we take in lots and lots of students and ... learn from them, what new technologies they're using."
Soderstrom also says that he's looking ahead to some longer-term technologies like quantum computing, but they aren't ready for the prototype phase yet.
You can hear Soderstrom talk about NASA JPL's prototyping process, and conduct a breakout session on how the organization used a multi-cloud model to support the Mars rover landing, at the CITE Conference and Expo on April 27 through 29 in San Francisco. Register here.