Eventos
Antonio Alfredo F. Loureiro
Federal University of Minas Gerais
Department of Computer Science
Belo Horizonte, MG, Brazil
Dr. Antonio Loureiro is a professor of Computer Science at the Federal University of Minas Gerais where he established probably one of the first research groups in wireless sensor networks and ubiquitous computing in Brazil. Currently he conducts several studies in those areas with his students and other collaborators in Brazil and abroad. In particular, Dr. Loureiro is a member of a national project in the area of sensor networks. Dr. Loureiro holds a PhD in Computer Science from the University of British Columbia, Canada.
Abstract
Wireless Sensor Networks in Environment Monitoring
The concern about environmental conservation and preservation assumes importance in proportion to the magnitude of tropical, sub-tropical and temperate forests available from Mexico to Patagonia. In particular, tropical forests in Central America as well in the Amazon region including Brazil, Bolivia, Colombia, Guyana, Peru and Venezuela have attracted much attention from different segments of national and international communities due to their size and diversity of flora and fauna. In October 2010, World Wild Fund for Nature (WWF) published the report “Amazônia Viva: Uma Década de Descobertas 1999-2009” (Amazon Alive: A Decade of Discoveries 1999-2009), which reports that between 1999 and 2009 scientists identified 1222 new species in the Amazon region – this means the discovery of a new specie every three days on average! Given the challenge of obtaining data from those forests to conserve and preserve them, a natural question arises: how can we use Information and Communication Technology (ICT) to help scientists and conservationists in that challenge? In the following, I discuss how a wireless sensor network can be used in that direction. A Wireless Sensor Network (WSN) consists of spatially distributed autonomous devices that cooperatively monitor physical or environmental conditions, such as temperature, sound, luminosity, vibration, pressure, motion and pollutants. The potential for observation and control of the real world allows WSNs to present themselves as a solution to a variety of monitoring and control applications, such as environment monitoring, biotechnology, agriculture, urban sensing, medical control, among others.
Traditional methods of sensor networking impose an impractical demand on cable installation, requiring wireless communication and motivating the use of WSNs to allow distributed processing. This is a new technology that allows the integration of the physical world to the Internet. A WSN is a sensing tool for collecting distributed events, processing and dissemination of acquired data and processed information to observers who are typically outside of the monitoring region.
Among the different species found in the nature, frogs are natural candidates to be used as indicators of environmental stress. Frog is an amphian, which represents a class of the animal kingdom highly sensitive to ecological disturbances both in aquatic and terrestrial environments. They demand specific requirements from the microhabitat where they live and have specialized physiological needs. Thus, the decline of the amphibian population can be used doubtless as an early indicator that there is some problem in a given ecosystem. Thus, besides the importance for the conservation and preservation of amphibians, their proper monitoring plays a very crucial role for the assessment of ecological problems in the early stages. In this case amphibians serve as a highly sensitive biological sensor of their environment. In particular, a monitoring tool for early detection of environmental disturbances is of great value to the preservation of the rainforest in Central and South Americas, where there is a great number of amphibians. Frogs are abundant, easily sampled, are stable in numbers and are particularly sensitive to environmental disturbances. However, to avoid an unnecessary environmental stress, the act of monitoring frogs in their habitat must be ideally the least intrusive. The detailed knowledge of parameters such as presence/absence, population density, dispersal and reproductive success of different species of frogs may allow a viable study for their preservation including their ecosystem. This would allow us to know better their social and reproductive behaviors; density of group population; study of territorial disputes between individuals/groups; eating habits (diet and schedules); correlation between climatic effects and their behavior; correlation between their population with the increase/decrease of other species (e.g., insects and mammals), and correlation between their population with the quality of the their habitat (e.g., water, air, temperature, humidity). Currently, the act of monitoring frogs demands the physical presence of a researcher to capture their audio or to listen to them, to make a proper classification. However, the simple presence of a human being interferes in their behavior and in the ecosystem to be studied. As frogs are very sensitive to such environmental changes, the human presence interferes with the acquisition of data and the analysis of the environmental impact to the ecosystem. WSNs emerge as a viable technological alternative, given their characteristics as discussed above.
There is a great opportunity in applying wireless sensor networks for monitoring frogs in those forests. Based solely on the sounds emitted by frogs, we can identify the different species present in the microhabitat, sex, and purpose of each vocalization. For instance, the frog “blue finger” found in the Amazon region has three different vocalizations. The first one is used to establish territorial boundaries (the sound seems whistles in continuous notes). The second one is used to deal with male intruders (the sound is low and hoarse). The third one is used to date a female frog and stimulate the oviposition (the sound is low and velvet). The rationale behind this project is as follows: there is no doubt about the importance of monitoring a rich and important environment such as the Amazon forest. Frogs serve as a highly sensitive biological sensor of their environment. In other words, they work as an indirect sensor of the environment where they live. Frogs are abundant, easily sampled (basically we need to collect audio and be able to detect their presence; sensor devices are placed in the environment and are not attached to the frogs), are stable in numbers and are particularly sensitive to environmental disturbances. On the other hand, the physical presence of a researcher to monitor them interferes in their behavior and in the ecosystem to be studied. Therefore, a wireless sensor network emerges as a viable technological solution not only to monitor frogs in a non-intrusive way but also to collect precious environmental data from their habitat that can be used by researchers to answer technical questions that do not depend on their presence. This opens the possibilities of studying problems even not thought before.
The deployment of wireless sensor networks in the Amazon rain forest will be also particularly important for Computer Science/Engineering researchers who will have a real testbed to study and better understand problems related to data communication and algorithms in WSNs. Furthermore, there will be professionals who can gain experience and knowledge in this area.