Technology Disappoints in Tsunami Relief
Human remains lay scattered amid rubble and the stench of decay in Banda Aceh, the region hardest hit by the Dec. 26 tsunami. Three Pentagon observers had come to study how well the U.S. military was communicating with humanitarian groups trying to feed and house millions of displaced Indonesians.
But the military's training, computer networks and procedures were tightly controlled. They were geared for sending helicopters, ships and soldiers to war—not for helping civilians in a disaster zone put a roof back over their heads.
More than two weeks had passed since tidal waves swept 300,000 people to their deaths in Thailand, Malaysia, Sri Lanka and eight other countries surrounding the Indian Ocean. About two-thirds of the dead and missing were in Indonesia's Aceh province, where homes on a 150-square-mile strip of coastline had been crushed, and many of those inside were drowned or washed out to sea. More than 500,000 survivors were homeless.
The expedition's leader was Cmdr. Eric Rasmussen, a Navy doctor who had participated in humanitarian operations following a 1999 earthquake in Izmit, Turkey, as well as in wars in Bosnia, Afghanistan and Iraq.
Even before he left Indonesia's capital, Jakarta, for Banda Aceh, Rasmussen had assembled a long list of logistical problems linked to communications breakdowns. On Jan. 9, in an e-mail to the United Nations, various humanitarian agencies and the acting Assistant Secretary of Defense, Linton Wells II, at the Pentagon in Washington, D.C., among others, he wrote:
Much of the work here is being done by cell phone, VHF radios and personal conversations. There is no Instant Message system, there is no collaborative space, and there is no consistent update of the information flowing around the theater. It's almost as if Third Fleet's Joint Operations Center had never existed.
Why was a doctor putting lack of a "collaborative space" so high on his list?
Because in its absence, aid workers were wasting time up in helicopters, surveying areas of the disaster zone that others had already assessed. With more sharing of information, they would already have been packing those helicopters full of food for the hungry and tents for the homeless.
Perhaps no one would die as a result—most fatalities occurred within minutes of the disaster—but that lack of "collaborative space" would cause needless misery for survivors, not to mention more wear and tear on relief workers.
Rasmussen had spent much of the past five years preparing for just such an event as this.
In 2000 and 2004, he had organized a series of exercises called "Strong Angel," designed specifically to address ways the U.S. military could better assist emergency humanitarian relief efforts in disasters worldwide. The second set took place on a bed of crushed lava in Hawaii meant to simulate the austere conditions that might be found in the Iraqi desert—or in the aftermath of a natural disaster.
The collaboration between military, academic and international relief organizations drew on diverse strains of influence, such as the Burning Man Festival, an annual arts gathering that coordinated itself through a high-speed wireless network in Nevada's Black Rock Desert, and the ready availability of "cheap and cheerful" technologies such as "cantennas," which could put a powerful directional antenna in anyone's hand for less than $50.
Rasmussen and colleagues inside and outside the military had been promoting the idea that the restoration of roads, bridges, ports, water, sanitation and other basic human needs could be planned for, practiced on and responded to quickly through the smart use of such cheap technology.
But here in Aceh, in a real crisis, Rasmussen saw relief efforts rendered inefficient by a lack of basic information— where the water, medicine, wood and other supplies were needed most, and how to get them there.
The scope of this series of tidal waves made the response particularly challenging.
Only a few other calamities—a 1970 cyclone in Bangladesh, a 1976 earthquake in China, and the famine in Ethiopia in 1984 and 1985—had cost more lives. The rush to offer assistance only complicated communications and coordination. More than 100 relief agencies, including CARE and the International Medical Corps, converged on the disaster area. Combined with military hospital units from many nations and an assortment of U.N. agencies, coherent management of the overall operation would prove close to impossible.
Observing along with Rasmussen and trying to suggest ways to make the mission work better were Dave Warner, a collaborator on Strong Angel and other projects, and Dan Engle, a retired Navy networks architect and independent consultant.
When not on a humanitarian mission, Rasmussen is a full-time doctor at a Navy hospital in Bremerton, Wash. But he also manages projects for the Defense Advanced Research Projects Agency (DARPA) and the Center for Robot Assisted Search and Rescue, which sent mechanical searchers into the ruins of the World Trade Center. He has taught medicine at the U.N. Office of the Coordination for Humanitarian Affairs in Geneva, and lectured on how to examine torture victims. The U.S. Agency for International Development (USAID) has him on the short list of people it calls to coordinate the humanitarian response to disasters around the world.
All of these roles—military man, doctor, humanitarian and technologist—surfaced in the Strong Angel exercises. In particular, the second set, held last July in Kona, Hawaii, emphasized technologies that could support electronic collaboration either on a battlefield or in a disaster zone. One ground rule: Participants had to rely solely on satellite and wireless communications gear they could carry.
The system they pieced together during the weeklong exercise included a Web site and a system of synchronizing folders—a "collaborative workspace" known as a Groove network—that could have sped up road clearing and the delivery of fresh water in Banda Aceh. That much was clear to Nigel Snoad, chief information officer of the U.N. Joint Logistics Centre (UNJLC) and a Strong Angel II participant.
Summoned back from vacation, Snoad had arrived in Indonesia on Dec. 28. He acted as a logistics coordinator, fielding calls asking basic questions like how to get a plane into Banda Aceh, how to get a car, and which roads were open.
At one point on Jan. 2, he shut off his cell phone for a half-hour to talk to a relief worker. When he switched it back on, he found 58 voice mails waiting. "It was driving me insane," he says.
If the tsunami mission could have deployed the kind of inexpensive but effective networking found on the lava beds of Strong Angel II, relief efforts in Aceh would have included:
- Voice, video and data networks that could be installed in any improvised office space—or tent.
- Wireless communications, linked to the Internet by satellite, and shared freely among military and civilian relief workers.
- "Virtual" workspaces for organizing deliveries of services and materials that non-technical relief teams could create, on the fly.
- Synchronization of documents that required only the occasional ability to get online.
- A "Pony Express" system of communications vehicles that could bring network access to users, as needed.
Rasmussen wasn't surprised to find that few of the technologies tested in July in Hawaii had found their way to Indonesia six months later. But he still hoped to prove austere networks could make a difference in Indonesia—and, in so doing, improve future disaster relief efforts.
In particular, he hoped to test tools developed with the help of software supplier Groove Networks that could wirelessly replicate databases wherever they might be found.
After all, what could be more important than matching offers of assistance from around the world to requests for help? Or finding the doctors, logistics experts or other specialists who had come to the scene to help, but were not hooking up with the people who could put their services to use?
But getting into the Groove products was not easy.
Shortly after arriving in Banda Aceh, Rasmussen found himself standing in deep mud at a makeshift airfield that had been created out of a drenched soccer field. Forget communicating from one computer to another. With helicopters roaring overhead, just carrying on a face-to-face conversation with Gregg Nakano, the head of the USAID Disaster Assistance Response Team (DART), was difficult.
The DART leader, who had worked with Rasmussen in Iraq, was charged with supervising the distribution of emergency aid, primarily through private charities like CARE and Save the Children. But he faced the same limitations: He was limited to cell phone communications—no access to Web sites, no e-mail, no radio communications.
He was scrounging for detailed maps and aerial photos that would show where the roads were still passable, and thus food and clean water could be sent by truck instead of helicopter.
In the Strong Angel model, any member of DART's relief team or any other relief organization would have maps or photos on their hard drive, for ready access.
With Groove, members of the network can see designated files on each other's laptop or desktop computers. When any member makes a change to any of the designated files, other members' copies of the files automatically update each time they connect to the network.
There are other ways to share files, like posting them to a public Web site. But in Aceh, the maps and photos were bottled up on servers aboard Navy ships such as the U.S.S. Lincoln, an aircraft carrier whose helicopters were being used to distribute food, medicine and other urgently needed goods. Only military personnel—offshore—had access to the most current information.
While on board ships operating off the coast of Aceh, Engle had been hearing similar complaints about lack of information sharing from Navy pilots.
Particularly in the first days of the mission, the pilots found themselves repeatedly flying the same "assessment missions" up and down the Indonesian coast for different relief organizations because the photos and observations brought back by one weren't shared with the next.
"While we were flying these assessment teams around for all these different organizations, we could have been flying food and water," Engle says.
In the meantime, Navy pilots swarmed the half-flooded soccer field, with seven helicopters doing the work of 30. Female soldiers, ankle-deep in mud, lugged 40-pound bags of rice, loaded them onto choppers, then turned around and did it again. And again. Only their muscles could speed up the relief effort, reducing each flight's turnaround time.
With a collaboration "space" and the network to support it, reports from observers could have been available to all.
"The problem was there was no such place, and no plan for that," Engle says.
At least Banda Aceh had a working cellular network, although it was frequently overloaded. Farther afield, where fishing villages had been wiped off the map and survivors huddled around campfires, even the cell phone wasn't an option.
Rasmussen instead had brought along a phone that operates as either a cell phone or a satellite phone, working with the orbiting relays of Thuraya Satellite Telecommunications of Abu Dhabi. In fact, he had brought extras, which he gave to the relief workers who needed them most.
That experience in Aceh reaffirmed for him that the "social network" between relief workers, doctors, government officials and military officers was more important than the electronic one.
Finding a single person like Snoad here, familiar to him from the Strong Angel exercise, let him make connections with other key players and get more work done than he could have otherwise, regardless of the technology at hand.
"Technology is not the point," says Rasmussen. "It is simply the facilitation technique."
A big reason he had come to Banda Aceh was to fill in the gaps of his wish list for information on what was happening on the ground; he had been collecting information from various relief organizations he met along the way.
There were no secrets here. There was information that could be, and should be, shared freely online—updated maps, satellite imagery, and aerial photos of the roads and likely areas where survivors might be found—but much of it was bottled up offshore, in the server rooms of Navy ships.
The reasons weren't entirely technical. Diplomacy often dictated that the U.S. release information to the Indonesian military rather than directly to relief organizations.
So, once Rasmussen managed to navigate the bureaucratic maze, he found himself acting as the highest-speed connection in this emergency network. He would fly out to the Abraham Lincoln, burn a CD with maps, photos and other key data on it, fly back to shore and share it.
Warner makes a techie joke of the experience. "What's the bandwidth of a kilo[gram]'s worth of CDs?" he asks. "Pretty good, but kind of bursty."
Strong Angel II had run on a free wireless network, connected by satellite to the Internet. Something like that would have been very useful in Banda Aceh. In fact, the French organization Telecoms Sans Frontiers (Telecommunications Without Borders) was trying to set it up, but the satellite connection it shared was limited to the bandwidth of a dial-up modem.
The U.N. was having its own problems establishing a high-bandwidth satellite connection.
One problem, Snoad says, was that communications kits for rapid deployment were optimized to work with the satellites over Africa, where the U.N. has seen the most frequent humanitarian emergencies in recent years. The kits were not adjusted for satellites hovering above the Indian Ocean.
Meanwhile, the U.S. had brought plenty of bandwidth ashore, but it was not the kind of bandwidth that could be shared with outsiders.
Limited to a single military transport plane for his gear, the Marine Corps officer setting up the onshore network brought communications equipment that supported the military's classified Internet protocol network, SIPRNet—which is only meant to create a secure, private network.
No one could fault him for following standard operating procedures, Engle says, since maximum security is the military's default mode. But one recommendation he made upon his return was that the military develop a different "fly-away kit" of network equipment for use in humanitarian operations, where information sharing is more important.
"Other than food and water, communications infrastructure should be one of the first things considered," he wrote in his preliminary report to the Pentagon on Feb. 16.
Back on Jan. 8, when the observer team stopped at the joint command center in Utapao, Thailand, en route to Indonesia, Engle was struck by the lack of what the military calls "situational awareness."
Generally, that refers to a military commander's understanding of what is happening on a battlefield, based on the information available at a particular moment. For example, the Navy uses its WebCOP software to give its crews a "common operational picture" from wherever they are, through a Web browser. The system constantly updates maps of a given region, superimposing locations of friendly and enemy forces, to help commanders make better decisions.
A humanitarian mission ideally would have a common operational picture of where food, water, transport planes and trucks are, and where they needed to go. What Engle instead saw at this military command center was a single PowerPoint slide. Projected on the wall was the status of available transport aircraft, among other items. But the data was static, not updated on the spot.
"I would have expected more, frankly," Engle says.
But the U.S. military's battlefield collaboration and decision-support tools aren't well suited to a humanitarian operation, where information needs to be shared broadly. Tools like WebCOP are designed for use in command centers with high-bandwidth networks, whereas the ideal situational awareness tool for a humanitarian operation would be accessible to team leaders in the field with only sketchy network access.
This was essentially what the Strong Angel team tried to construct in Groove Virtual Office, and which Rasmussen and his collaborators from Groove transformed into a virtual emergency operations center after the tsunami.
Groove was founded by Ray Ozzie, best known for creating Lotus Notes. Ozzie, now one of Microsoft's three chief technology officers after his company was acquired in April, wanted to create a new kind of collaboration software for workers from many different organizations, who might spontaneously form teams to complete specific tasks.
Rather than relying on technology staff to set up databases and file servers for shared work, Groove users can establish "workspaces" on their own personal computers and invite others to participate.
A workspace for a given project can be assembled easily from templates for sharing information and files, or it can be custom built.
Data is replicated directly on personal computers. Servers in a network play only a coordinating role. To save bandwidth, only changes to a document are transmitted over the network. In the process, the application encrypts (and decrypts) information so it can operate securely on public airwaves or land lines. No firewalls needed.
When Rasmussen discovered Groove, he saw a tool for bridging the divide between military and civilian participants in a humanitarian mission. He convinced Ozzie that his company should help adapt it for that purpose. Because it wouldn't be running on any one organization's servers, a Groove application would put everyone on a level playing field, Rasmussen reasoned. That would make civilian organizations more likely to participate than if the U.S. military owned and operated the system. Meanwhile, Groove's encryption made it more acceptable on his side of the divide.
Rasmussen brought Groove's software with him to the military's joint command center in Tampa, Fla., in November 2002, when he was helping organize a humanitarian response to the coming U.S. invasion of Iraq. He carried it with him into Kuwait and then Iraq in the spring of 2003.
Working over a long-distance hookup and across time zones with Groove developers, he came up with an assessment form to measure whether the Iraqi people's quality of life, compared with international norms for health and nutrition and other measures, was rising or falling. The form could be shared and synchronized by relief organizers after the U.S. invasion.
He then gave the form away to humanitarian workers who could use it, encouraging them to download Groove's free trial version.
The Strong Angel II team developed more Groove applications for humanitarian use, including an early version of databases for tracking requests for help and offers of assistance. Following the tsunami, Rasmussen asked his friends at Groove to assemble those tools and any others that might be useful into one package that could assist the relief effort.
Rasmussen got a chance to demonstrate the usefulness of Groove shortly after he arrived in Jakarta. He checked in at the U.N. command post, met friends from the World Food Programme, and heard about the trouble they were having establishing a satellite link to Aceh.
Rasmussen popped open his laptop and saw a green icon next to the name of Enrica Porcari in his Groove contact list.
That meant Porcari, who was both a Strong Angel participant and formerly a WFP field director for telecommunications and emergency operations, was online. Better yet, she was close by, working in Malaysia, just across the Strait of Malacca from Aceh.
Though she was now CIO of another U.N. agency specializing in sustainable agriculture, the Consultative Group on International Agricultural Research, she knew emergencies, knew the region, and knew telecommunications. The gist: She had a lot of expertise to offer the WFP in Jakarta, and yet the food relief agency had no idea where she could be reached.
Rasmussen sent an instant message, got her cell phone number, and within minutes had recruited her to help figure out both the satellite networking problems and more basic practicalities like locating translators who knew the languages spoken in Aceh.
It took Rasmussen longer to find someone at the Jakarta command center who could sit still long enough for a full demo of the Groove application. Finally, he pinned down a U.N. logistician and gave him the guided tour. But before he was finished, his victim's eyes glazed over.
"He finally said to me, 'Uh, now there's Groove under here somewhere, right?'" Rasmussen recalls.
Looking at the application anew, Rasmussen realized he had pushed the Groove developers to layer on too many menu items and tabbed windows. The easy-to-use features of Groove, such as directories of who belonged to a given Groove network, were buried. The whole thing had to be radically simplified, he realized, if it was to get used.
Making apologies, Rasmussen retreated to his hotel room and woke up Robert Kirkpatrick, Groove's humanitarian solutions architect, in Boston, 12 time zones away.
They talked through the changes on Skype, a service that allows calls to be made around the globe to other Skype users. All he needed was a headset and some voice-over-Internet Protocol software on his laptop.
When Kirkpatrick and Rasmussen were finished with their redesign session, they had stripped away everything that was nonessential, trimming the table of contents that greets users from 28 links to five major categories—Discussion, Contacts, News, Files and Toolbox—with subheadings for the specific tools created for logistics support and field assessments.
By eliminating clutter, they found room for some explanatory text at the top of the page about how to use the member directory and online collaboration features. The basic feature that allowed Rasmussen to wow his WFP friends when he first arrived in Jakarta—the ability to connect people to each other—was the most important of all, they decided.
Before departing for Banda Aceh the next day, Rasmussen took the revision back to the U.N. logistics specialist he had tried to train and asked him to take another look. After that, it did get some operational use. But it was late.
In retrospect, he wishes he could have gone to Indonesia sooner, perhaps allowing him to introduce Strong Angel-style capabilities while the mission was still being organized.
Snoad, the U.N. Joint Logistics Centre CIO, confirms that Rasmussen's Groove tools were used in the tsunami response, "although not to anything like their proper potential."
Engle says he thinks Groove-like sharing makes most sense as a supplemental tool in an emergency, for occasionally connected users, rather than as the main way to maintain "situational awareness" during a massive humanitarian response. His laptop, for instance, kept crashing when he tried to access the tsunami response workspace because Rasmussen had packed too much into it, he says.
Whatever the technology, the right time to introduce such capabilities is during the planning for future emergencies, Snoad says. The point is "getting the international community to be more organized and professional, not just about the tools but the approaches as well—and recognizing that it matters."
Within the U.S. military, such recognition is growing. In Afghanistan and Iraq, military invasions were followed by reconstruction missions involving the Pentagon, humanitarian relief agencies and the United Nations, designed to convince the people of those countries that the U.S. wasn't hostile to them, only to their former leaders.
Even before those conflicts, the military was participating in humanitarian missions with increasing frequency, says retired Vice Admiral Dennis McGinn, who instigated the first Strong Angel exercise back when he was commander of the Third Fleet. He kept championing the idea that the military needed to practice for humanitarian operations, just as it did for war, when he went to the Pentagon as Deputy Chief of Naval Operations, in charge of warfare requirements.
"It's gotten better over the years, but it's far from where it needs to be," McGinn says.
In the 2000 Strong Angel exercise, Hawaiian citizens were recruited to play the part of refugees, and participants in the exercise practiced caring for them while simultaneously watching for terrorists in their midst. Rasmussen had the contacts to pull together a diverse group of participants from the World Food Programme, other U.N. agencies, and charities such as the Red Cross, as well as the militaries of Australia, Japan and other nations.
In 2004, Rasmussen brought a similar group of international organizations together with the U.S. Army, Navy and Marine Corps and the Coalition Provisional Authority for Strong Angel II.
Groove played a major role in that exercise because Rasmussen planned to make its software the centerpiece of the Strong Angel collaboration system, integrating it with a variety of other technologies, including the following:
- TrIM, translation software for instant messaging developed by Mitre, a federally funded research lab. A version that worked in 24 languages was integrated into Groove's instant messaging system to create "Babylon Chat." Among other things, English- and Arabic-speakers could type and receive messages in their own language.
- Geographic positioning technology, which could immediately identify on a master map the location of every participant in an exercise who called in with coordinates.
- Low-bandwidth videoconferencing, using software from San Jose-based startup Vsee Lab.
Rasmussen and Warner also turned to technician Clif Cox, known for his expertise at setting up high-speed wireless networks at the Burning Man Arts Festival, known for "radical self-expression" symbolized by the torching of a giant wicker man. Cox and cohorts set up camp in Nevada's Black Rock Desert to operate what Rasmussen calls "the best austere environment communications network you will ever see."
Among other things, the Burning Man crew extended the range of the camp's WiFi network by taking advantage of the local landscape and setting up a transmitter on top of the Kauhola Point lighthouse.
The combination of Groove and wireless networking led to the creation of the Strong Angel "Pony Express." To make the network truly mobile, they drove around the island in a Chevy Blazer dispensing doses of connectivity to a remote spot with a directional antenna. Specifically, they used a cantenna, a simple WiFi directional antenna originally designed by hobbyists who used a foil-lined Pringles potato-chip can to focus the radio beam. In place of a junk-food container, they used a commercialized version, the Super Cantenna, which costs about $50.
These connections are designed only to last a few minutes, so the key to making the Pony Express work was the Groove Relay Server loaded on a laptop in the back seat of the Blazer.
The relay server, which normally runs in a data center, is the part of the Groove system responsible for storing temporary copies of updates to a workspace. Without it, the updates might not be available when other members logged in to receive them.
By the time the Pony Express would complete a circuit of remote locales on the island and return to the Strong Angel base camp, it had sent and received updates from all the otherwise disconnected users in the exercise.
Once it finished updates from workers at the base camp as well as people able to use the camp's Internet connection to synchronize information with workspace participants in other physical locations, the Pony Express was ready to start its rounds again.
Something like the Pony Express could have been useful in Indonesia, not so much for those in Banda Aceh as for the humanitarian workers farther out in the field. With roads often impassable, the delivery vehicle might have been a helicopter, but the general concept is the same. "There's a lot of interest in this right now," Rasmussen says.
The DOD's Wells says Rasmussen's approach is gaining currency within the Pentagon, which continues to advance "network-centric warfare" that provides better information to commanders in the field, as action takes place.
The Strong Angel approach is more radical, though, because it suggests extending the decision-making network beyond the military's own boundaries to include other organizations, such as humanitarian relief agencies. That would require changes to the codified set of rules that govern military operations, he says.
But recent events might make that more plausible. "The combination of the Iraq experience and Afghanistan suggests to a lot of people that the time is right" for making communications a fundamental part of emergency responses worldwide, Wells says.
The approach Rasmussen has been promoting "is now coming to be recognized as a valid mission," he says. "Instead of a few folks laboring in the wilderness, it's being recognized as a valid, operational thing" to consider cheap, easy-to-use communications as much a fundamental requirement of disaster relief as bags of food or carts of medicine.
Strong Angel Team: Basic Stats
Organization: A virtual team that include military, medical, humanitarian and technology experts.
Web Address: www.strongangel.telascience.org
Business: Define better ways of using communications technology to support collaboration between military and humanitarian organizations during emergencies.
Project Manager: U.S. Navy Cmdr. Eric Rasmussen, M.D.
Budget in 2004: About $330,000 from the Defense Advanced Research Projects Agency.
Challenge: Identify computer and communications technologies that can withstand harsh conditions—in Iraq or the aftermath of a tsunami. Push them from experimental to practical usage.
- Increase number of significant relief agencies that can tie into Strong Angel network, from 20 in 2004 to 50 or more in 2007.
- Boost use of electronic collaboration tools, from 13% of relief organizations after December 2004 tsunami to 50% during next emergency response.
- Increase number of humanitarian specialists in Strong Angel online contact list, from 175 today to 500 by 2007.
- Convert Strong Angel project team from 10 part-time workers in 2005 to permanent organization with full-time staff by 2007.