DigitalGlobe has been hard at work setting up their Analysis Center to illustrate the utility of their imagery, as well as to sell their capability as a service. Matt Ball recently spoke with Stephen Wood, vice president of DigitalGlobe’s Analysis Center, and Chuck Herring, director, about their capabilities as well as the process of turning their information into a service.
Ball: Despite all the technical advancements with satellites and imagery delivery, I can’t help feeling that we’re still just scratching the surface in what we can do with imagery, particularly with the addition of your 8-band multispectral imagery, and upcoming sensors.
Herring: I remember one of the first media interviews that I did back in the EarthWatch days where the reporter lectured me that taking an image meant that as soon as you captured an area you were done. Now with the ubiquity and the use for so many different application areas, everyone understands that you’re just done for that day. Customers have progressed from wanting imagery once a year to once a month, and now many customers are eager to get imagery as soon as possible.
Wood: I think we do a great job of selling and explaining within the geospatial industry, but the challenge and part of the catalyst for the Analysis Center team here is to explain it to those that aren’t familiar. How do you take an 8-band image that a remote sensing spectral analyst is excited about and explain it to someone at Morgan Stanley? They don’t care about the spectral bands, but they could take the information derived from the image if it is presented in a meaningful way.
Part of what we’ve been trying to do is to be that bridge between the technical community and the business community. We present information in simple to understand ways, where you don’t need to be a remote sensing or GIS geek, but with all that power behind us.
Ball: Where have you invested the most in terms of your R&D budget?
Wood: I often like to point out to people what we’ve done on the ground, even though it’s not as glamorous and as sexy as the satellites. We have a phrase that got us through a big bid, “we have to do things equal to every satellite imaging company in space, but we have to win it on the ground.” If you can’t get the data to the ground in a meaningful and timely way, then you’re missing a big part of the story.
We’ve focused a lot on speed, and the ability to take all this rich content and distribute it around the world. Not only was it millions of dollars and years spent here at DigitalGlobe, but it often gets overlooked because we’re so proud of the satellites.
IT people love to hear about servers with petabytes of storage, but most people want to see the satellites. We fall under that trap very easily, but it’s equally important to think about how the imagery gets to people.
Herring: Now, we’re seeing imagery within an hour of being taken. It really makes a difference, especially when we get pushed by our customers to deliver imagery quickly, such as with natural disasters. Waiting two days to see an image is not meaningful anymore.
Ball: Is your historical archive also a big part of what’s meaningful for the Analysis Center?
Wood: The extent of coverage and the constellation that we have provides a time machine of the world. We can literally go back to any spot on the earth over ten years and see what has changed. We have been steadily building our image library and we’re close to 2.5 billion square kilometers in this library. Part of what has enabled our team is our direct access to that library.
We’ll get an inquiry from any organization looking to see what has changed, so they can perhaps understand what will happen in the future. My team can pull in this treasure trove and mine it, and then I can go to collection and planning and get a shot for tomorrow. To me that is unique, because we have both the foundational depth as well as the agility to respond to current events. To really advance remote sensing and its ability, you have to have both. You can’t just do one or the other.
At the macro level, this industry is all about change, whether that’s business, agriculture/forestry or military operations. Seeing what has changed is the nature of the geospatial industry.
Herring: Even the people that aren’t imagery savvy want to see what has changed. That’s how Jane Goodall approached us years ago when she realized we could help. She had people on the ground, but understood that she needed to monitor the change in Gombe in a different way. Even non-imagery people understand the need to document and see the change, as well as monitor that change. Speaking to that change doesn’t do it but, showing an image puts the change in context.
Ball: I recently listened to a radio program that was talking about the advances in DNA evidence, where the smallest trace is enough to provide evidence of crime. Drawing a parallel, we’re amassing imagery that can become evidence in terms of planetary change, but we need advances in how we analyze it.
Wood: I think that’s a challenge for the remote sensing imagery as a whole. How do you take 2.5 billion square kilometers and mine it in an expeditious way? I’ve been looking at satellite imagery for 25 years, and each year there has been hope that there would be an automated solution for feature extraction or change detection. We’ve been waiting for the scientists to hand us an application and we’re still waiting. My challenge to our R&D team is to take a step back from just the science to figure out how we can do things faster for our customers.
We use all this incredible technology for purpose. We don’t need to always be looking for the super glamorous solutions. Sometimes it can just be stitching together imagery of an area in a historically-accurate way. The human mind and the visual eye can see things that are changing.
Ball: How do you go about marketing change as a service?
Wood: We’ve gone through eras of resolution, accuracy and speed, and now we can help world awareness. Now we spend a lot of our time responding quickly to crisis events. Sometimes we’re the only people that can supply actionable information.
We’re empowered by all kinds of sensational technology, but it all comes down to the people. One of our analysts found the Costa Concordia as it ran aground. People who impart wisdom can quickly tell a story.
With the Arab Spring, we were on the phone with news producers, but also getting calls from financial firms worried about oil. They wanted to track ships going up and down the Suez Canal to see if there was a decline in the shipping of oil in light of the political unrest.
With this view into change, business people have started to change their thinking to understand the application of this technology.
Ball: Does the FirstWatch service involve subscriptions to a specific view of a country or region?
Wood: It may be a specific view, but it can also be customized for an area of interest or to monitor a subject matter. What’s probably more indicative of FirstLook or FirstWatch is what we did in Japan. We were driving to see how fast we could respond. Part of it was out of a feeling that we wanted to do anything we could to help, and part of it was to see what we could do. Our team was here all night every night because as the imagery was collected it was the middle of our night here, it was the morning in Japan.
As the data would come down, we would process it, exploit it, analyze it, and then report on it. Our fastest response was two and a half hours from the moment of collection to the finished report going out. These reports would be pretty quick and simple but also pretty powerful.
We were watching the Fukushima Daiichi plant, and one day in particular on March 14, it was amazing that we could see Reactor 3 about 30 seconds before it exploded, and also go back three minutes later and do it again. We were able to take this data and show the smoke and dust plume as it was going out. We sent it to the Japanese government, the U.S. government, and the media in order to tell the story within minutes and hours that ten years ago would have taken us a couple of days to do.
Ball: So, the second shot was captured by a different satellite?
Wood: We captured the image, and then our next satellite was there 30 seconds later to capture the same spot. People always challenge us and say that we have to be in the right place to be able to do that. It’s true that physics are physics, but you also have to have the right number of assets to be in the right place. We’ve had multiple examples over the years where several of our satellites may be over an area where it is too cloudy to see, and then QuickBird comes in at a higher off-nadir angle and gets a perfectly clear shot.
Ball: The implications for that global view are very interesting from an energy perspective, having talked already about both oil and nuclear energy.
Wood: The Japanese have shut down all their nuclear power plants because of what happened at Fukushima. From a remote sensing or analytic perspective, you start to think about the Japanese dependence on other sources of energy, whether it is oil refineries or coal.
We’ve spoken to a number of companies that want to have us start tracking coal and the implications on countries like Japan’s economy. Can you use an indication such as the amount of coal that is being stockpiled as an economic gauge? Those are all the derivatives that we see coming out of crisis events. We start with the humanitarian angle and over time more business outcomes are coming out of it.
Ball: That eye on the news, and ability to respond rapidly, really creates an event archive.
Wood: The other thing we emphasize is the ability of satellite imagery to capture areas that are otherwise hard to capture any information about. While you’re looking at an area in the developed world where you have a glut of information from cell phones or aerial photography, try going into the middle of Libya during the NATO bombings and get meaningful information about the oil industry. If you want to get back into an area after the bombings have stopped, how do you do that? I think that’s where satellite imagery becomes invaluable, along with the ability to repeat visit and get the high quality data out. It’s hard to do that in any other cost effective way. That’s part of our daily operations now.
Ball: Is it getting easier to convey business value?
Wood: Another area that is growing is the ability to use these tools for competitive analysis. We’re integrating things that have been part of the domain of the defense world for the business world. The WorldView-2 8-band imagery is an incredibly technically powerful tool, but it’s a challenge to convey that to people in a meaningful way. We’re always working to convey information succinctly and effectively.
I was talking to Wall Street about FirstLook with the historical and current information, and they became fascinated by the industry, such as the floods in Indonesia and the hard disk manufacturers that were no longer operational. That’s the kind of information most clients are interested in because finance companies and investment companies are only getting the information that authorities want them to get.
Of course, they really wanted us to go into the building, and let them see the guy when he pushed the button to start the assembly line again. I had to manage expectations and let them know we could track the water as it recedes. I could see the heavy machinery coming in to clean up the place, and I could see the cars when the factory workers started showing up. You start getting into this pattern where you’re predicting what is going to happen.
Their eyes just open up as many see what we can provide. We’re trying to make the things that we’ve taken for granted in the remote sensing and GIS industry and get beyond that. We want to remain true to the industry because we have quite a history and pedigree, but the foundation was built to start taking this technology elsewhere.
Herring: We always focus on the human aspect, and we have started to look further around to understand the implications to the infrastructure, the industry, and the things that other people care about. We obviously want to help save lives, to aid humanitarian efforts, but what other things can be told, validated or denied?
Ball: Have you changed your outreach approach now that you have the Analysis Center?
Wood: A big part of what we’ve done this year is to both innovate and speak to customers, and break us away from our own line of thinking. It’s like the Wall Street story where we’d deliver a full report and get feedback that they want it on one page because of the pace at which they do business. We’re at a transition of going from a traditional way to a new, creative and innovative way of taking the tools that our industry has been using for 50 or so years and now applying them in a different fashion. That’s been challenging, but exciting as well.
Herring: I remember in the early inception of commercial satellite imagery value-added companies would not talk about delivering imagery but delivering answers. We’re starting to see that because the customers are asking for answers, and they don’t necessarily need the imagery.
Wood: That’s been a bit hard for us because we always want to talk about and show the imagery. We have several customers today that only want a statistical readout, because that is what they use to feed the rest of their process. That has been a real change and something that I think will continue to grow.
Ball: Have there been new applications or analysis work that has been particularly satisfying?
Wood: The work on the Arab Spring that we were doing in February and March (2011) was both personally and professionally satisfying. As the events were unfolding in Syria (2012), we were able to take our full asset source and capture it all. In a short period of time we collected more than a million square kilometers of data. We had great coverage in Homs, Hamah and Damascus that were being bombarded every day by the Syrian military. This was being displayed on the world stage, using our imagery, in a way that hadn’t been done before.
Part of this intersection harnesses all the Internet technologies and social media that are exploding today. We show such relevant data as human atrocities, and we’re able to convey a military precision within areas that are bombed. The human aspect that we were being asked to look for were mosques, universities, schools and hospitals. We had a sensational shot of an oil pipeline that was intentionally damaged, as well as an image that captured the smoke ring from a Syrian military gun that was shelling a village.
As a result in the success of getting that story out, we have worked more with the news media than we’ve ever done before. It’s partly because we have all the ingredients together, and we’ve made the business decision to highlight how this technology can help them. The Syrian crisis was a classic example because we could collect and analyze the images with a quick turnaround rate. It helped explain exactly what was happening, and it’s a dynamic that didn’t exist ten years ago.
Ball: What exciting technology is forthcoming with your next satellite?
Wood: The Worldview-3 satellite with SWIR really opens up new areas. I think WorldView-3 will extend our ability to extract information that is technically oriented. Our challenge will be to explain the importance of this technical trend. We’re confident of the kinds of information that it gives us on vegetation, land use and land cover. It gives us new capabilities for extractive industries such as mining and oil as well as military customers. Scientists that we’ve been working with are really excited about what they can do with the imagery. We’re now about a year and a half away from the launch.
Ball: The pace of digital mapping continues to accelerate. I remember a decade ago sitting down with National Geographic and they were so excited about pulling together a map of Afghanistan in just one month. And here you are sharing images of what’s happening on the ground on the other side of the world as it’s happening.
Wood: We’ve been working with several companies to take all our tools and the data and imagery that we have to map an area in just a matter of a day or a weekend. What we’re not going to do is have a team of several thousand people to do a map or to look at and exploit an image. We want to have ten thousand people helping us.
There have been a lot of discussions about crowdsourcing, and we’re fully involved in that right now. We’ve done some experiments using our web distribution technologies along with our remote sensing data and our analysis team to harness communities to exploit massive amounts of imagery in a short period of time. The vision is going in that direction, and using that combination of web technologies and people power with our core assets, gets to be very explosive. That’s something I’m really excited about for our future.