Advertisement
News
Advertisement

Crowd-Sourcing Helps Monitor Japan's Radiation

Fri, 05/17/2013 - 7:00am
Univ. of Southampton

A team of researchers from the Univ. of Southampton has designed a new tool to intelligently combine nuclear radioactivity data in Japan. The technology harnesses the power of crowd-sourced radiation data; an innovative resource that became available after the 2011 Fukushima nuclear disaster.

During March 2011, the second-largest nuclear emergency since Chernobyl 1986 was caused by a magnitude nine Tsunami hitting the North-East coast of Japan and severely damaging the nuclear power plant of Fukushima-Daiichi. The consequent nuclear accident provoked radioactivity increases of up to 1,000 times the normal levels in the area of Fukushima with more than 488,000 people being evacuated from their homes for the risk of nuclear contamination.

In response, private individuals brought forward the unprecedented effort of deploying 577 Geiger counters across the country to help the public monitor the spread of the nuclear cloud. These sensors were mostly built using low-cost open hardware boards such as Arduino and were able to stream radiation data in real time connected through the Cosm web platform. This crowd-sourced sensor network, also known as the Cosm network, came to life in less than two weeks after the Tsunami and provided very relevant data to both official authorities and local citizens for monitoring the evolution of the disaster. More recently, the network was extended to 1,024 sensors contributed by several other organizations such as Safecast and the Ministry of Education, Culture, Sports, Science and Technology of Japan (MEXT). All together, the Cosm sensors provided more than 27 million readings since the day of the Fukushima disaster.

According to the researchers, a key element in order to incentivize people to take part in crowd-sourcing projects is to help them understand these large quantities of data. To help people gain such an understanding, it is important to close the loop and feedback the results to the data contributors.

For this reason, the researchers have developed the Japan Nuclear Crowd Map (JNCM): a web platform that combines into a single database the sensor readings provided from the three main crowd-sourced radiation monitoring services: Cosm, Safecast, MEXT.

Matteo Venanzi, from the Univ. of Southampton, who developed JNCM says, “The platform automatically collects raw radiation data from the online sensors and uses a non-parametric Gaussian process model to fuse the data into a single radiation map over Japan. The estimates are then shown to the users as a heat map and an intensity map, showing the average radioactivity in each prefecture. The users can also search by postcode to find out the radioactivity in their neighborhood based on the latest predictions.”

JNCM is also available for smartphones with the JNCM Android app. Through the app, the users can visualize the radiation heat map directly on their phones as data are collected and also know the radiation level at their current location.

Yuki Ikumo, also from the Univ. of Southampton, who developed the JNCM Android app says, “JNCM aims to be one of the future technologies for disaster managements in which the large participation of people will play a crucial role in community-based sensing crowd-sourcing environmental monitoring tasks.”

JNCM users can now perceive the usefulness of this technology by freely accessing a number of radiation monitoring services based on the data contributed by thousands of crowd members.

JNCM is developed within the ORCHID project based in Electronics and Computer Science at the Univ.of Southampton, which investigates how human and software agents can work effectively together to collect the best possible information from a disaster environment. To find out more and try the platform, go here or download the app from Google play.

Advertisement

Share this Story

X
You may login with either your assigned username or your e-mail address.
The password field is case sensitive.
Loading