Ad- hoc Networks of Autonomous Sensors for the Management of Low Probability High Consequences Events

Early availability of reliable information is critical for optimizing emergency response and management in a disaster or crisis situations. Especially in cases involving search and rescue, it results in a key factor for saving human lives. Numerous tools that have been developed to support decision making in emergency management are centrally administrated and of a top-down approach. They usually have high setting up costs and need continuous maintenance in order to be kept functional and functioning at all times. Moreover, the substantial data collection and management activities on a just in case basis can result in important privacy related issues. In cases of low probability / high consequences events such as earthquakes or terrorist attacks an alternative / complementary approach is needed. Such an approach that overcomes, to a great extent, the above mentioned issues is based on the use of autonomous, sensor-based agents that have the capability to sense their environment and, on the occurrence of a certain event, react, individually or collectively. Such a prototype system has been developed to support search and rescue (SAR) operations in the event of a building collapse. It aims in providing quickly and cost-effectively reliable information, primarily on the occupancy of each space of the building during or immediately prior to (few minutes before) its collapse as well as on the building itself (i.e. number of stories, plan of each storey). The system is based on small, low cost, autonomous devices that, pre-installed in every building space, are monitoring continuously the structural elements, without storing or transmitting any information. Only when triggered by a collapse (i.e. significant change of the monitored structural elements) they store the sequence of images covering the last few minutes and then remain inactive in the rubbles. Upon arriving, the SAR team, equipped with suitable central units, wake-up the devices buried in the rubble so they form a dynamic network to upload their data. Processing and visualization of the uploaded image sequences enables a quick estimation and location of the trapped persons and, through audio link, communication with the eventual survivors. The use of similar systems, based on the principle of autonomous surveillance, has been proposed in public transport applications (i.e. explosions in trains) or for monitoring landslide risk.

ANDRITSOS Fivos;  DANDOULAKI Miranta; 

2010-07-01

Office for Official Publications of the European Communities

JRC32650

https://publications.jrc.ec.europa.eu/repository/handle/JRC32650,