Share |

The role of e-infrastructures in natural-disaster response

Feature | March 6, 2013 | By Andrew Purcell

A helicopter flies over the port of Sendai to deliver food, donated by the citizens of Ebina City, to survivors of the earthquake and tsunami.

Monday 11 March, 2013, marks the second anniversary of the most powerful earthquake in Japanese history, measuring a staggering 9.0 on the moment magnitude scale. The tsunami caused by this earthquake killed almost 20,000 people, and triggered a meltdown at the Fukushima Daiichi nuclear plant.

Ben Katsumi of the Japanese Information Technology Promotion Agency (IPA) believes that there are important lessons to be learned from the crisis regarding how e-infrastructures can best be put to use in response to natural disasters. He spoke about the role cloud computing played in aiding relief efforts in the immediate aftermath of the 2011 crisis at last week’s Cloudscape V conference in Brussels.

Following the 2011 earthquake and subsequent tsunami, hundreds of thousands of people lost their housing, daily lives, and jobs. The tsunami also destroyed social infrastructures, as well as IT facilities. People left homeless were evacuated to temporary accommodation in city halls, sports centers, and other public buildings, as well as temples, shrines and churches. Despite the swift relief effort, with huge response teams and armies of volunteers working together to help the victims, there was an immediate lack of food, water, medicine, sanitation, fuel and blankets in many areas. Relief teams were faced with the problem of not knowing exactly which resources were most needed where, as up-to-date information from those directly affected by the natural disaster was not always immediately available.

Earlier this year, scientists from the Japanese Atomic Energy Agency (JAEA) published research aimed at improving the clean-up of radioactive cesium leaked into the environment through incidents similar to the one that occurred at the Fukushima Daiichi nuclear power plant in 2011.

Zeolites, a type of microporous aluminosilicate minerals, are often used in water purification in a variety of industries — they’re also commonly used in cat litter. However, following the incident at the Fukushima Daiichi nuclear power plant, they were put to use cleaning up radioactive cesium leaked into the environment: sandbags of zeolite were dropped into the seawater near the power plant.

Zeolites are commonly used for radioactive clean-up purposes, since radioactive cesium isotopes readily adsorb onto their surfaces. However, many different specific types of zeolites exist and some are better at adsorbing radioactive cesium than others. The researchers at the JAEA ran calculations on a BX900 system at the Center for Computational Science and e-Systems to work out how to develop zeolites which were even better at selectively adsorbing radioactive cesium isotopes. They discovered that zeolites which have micropores with a diameter of around 0.6 nanometers, contain a moderate aluminum-to-silicon ratio, and have a uniform distribution of aluminum atoms around each of their micropores are generally much better suited to cleaning-up radioactive cesium.

The research, published in the Journal of the Physical Society of Japan, has the potential to significantly improve nuclear clean-up operations.

“This meant that effective and timely delivery of much-needed relief was impossible,” argues Katsumi in his paper for the Cloudscape event. “What was missing was information and the IT to convey, aggregate, disseminate and match information regarding needs and supply.” Katsumi goes on to point out that computers and other hardware items were also lost in the tsunami, with communication lines and servers destroyed, too. In order to solve this issue, cloud services were provided free to city staff and volunteer teams and cloud-based communication media also supported people’s peer to peer communication, allowing people find out if their family, relatives, friends and colleagues were safe. For example, the sinsai.info site was created using Ushahidi, a crowd-based incident reporting and mapping platform, and is managed under the Open Street Map foundation of Japan. This volunteer-run site, which was created just a mere four hours after the earthquake struck, used location-tagged information from people’s tweets to build up an accurate picture of the situation as the crisis developed following the natural disaster. Initially, the site was set up by a small group of the IT engineers lead by Haruyuki Seki, though many more e-volunteers soon joined, communicating with one another through online chat clients. The people behind the project, subsequently adapted the mapping tool for use following the 2011 earthquake in Christchurch, New Zealand.

In addition, cloud-computing infrastructure also helped local governments address a lack of capacity. “Many local governments had used poor [low-capacity] servers and narrow bandwidth to disseminate administrative information to citizens,” explains Katsumi. “The aftermath of the disaster saw demand increase to a spike peak hundreds of times higher as people clambered to get information on radiation levels.” Servers inevitably crashed as citizens searched online for the latest updates on the developing crisis. However, in response to this, many cloud service providers offered free hosting a mirroring to local and central governments, as well as some companies whose local IT infrastructures had been hit hard by the disaster.

“The disaster proved the cloud’s ability, efficiency and advantages in emergency response on a national basis,” Katsumi concludes. “Cloud is also used in every component of society and economy today. It supports people, companies, critical infrastructure and governments.” If clouds stop, the whole of society is affected, argues Katsumi. If this failure is due to a natural disaster, he warns that it could have the potential to throw both society and the economy into chaos and catastrophe. Consequently, Katsumi believes that it is of paramount importance to identify the potential damage that can be caused if cloud resources were to go down in a given region and to ensure that data centers are both robust and disaster ready.

Another important consideration according to Katsumi is the ability to move clouds from one data center to another. While he points out that virtualization technology makes this easier compared to traditional real machine-based computing, there is also a lot of work going on to develop standards to realize migration between clouds. However, to make cloud migration feasible, there are a number of technological, economic, and legal aspects which must first be addressed, argues Katsumi. These include: service-level agreements, security, consumer contract, provider-provider contract, compliance, and international alignment.

One of the most important factors in ensuring that cloud-based resources remained accessible during the crisis-response period was the fact that the data centers supporting the clouds remained online, says Katsumi. "The power network in Japan has lots of built-in redundancy, so even if several power plants go down, data centers should still be able to remain online." This, he says, was key to the role cloud-based resources were able to play in aiding the relief effort.