|dc.description.abstract||This paper presents a prototype of a system for automated observations of flora and fauna in the Arctic.
Currently applied methods of observation depend mostly on systems (usually consisting of a camera unit, a motion detection sensor and a memory card) that are left unattended in remote locations during extended periods of data gathering. The main problem with such approach is that no remote control or monitoring is available for those systems and manual inspection on site is not performed as often as it would be required for ensuring continuous operation. If a system fails, there is no way of detecting it, let alone fixing the issue or performing a reboot. Exposed to challenging environmental conditions of the Arctic and prone to problems such as power loss, hardware malfunction or inappropriate initial configuration, the systems have high probability of failing without it being noticed. In such cases, all several-months worth of data might be lost or never even recorded.
The solution presented in this paper intends to address the above issues by extending the functionality of an observation system with long range communication, self-monitoring and power saving capabilities. Proposed architecture allows for constant monitoring of system's health status and reporting it, together with sensor readings, via a remote gateway to the backend application. The system's designed uses IoT modules, which give it good extensibility properties if need for incorporating additional sensor types arises. The paper describes also the prototype implementation and the results of experiments performed.
The main focus of system test scenarios was on energy consumption, efficiency of data gathering and wireless communication capabilities. Currently the most serious concern identified for the system is its high energy demand. Experiments with different approaches to reducing the energy demand were conducted and presented in this paper. A satisfactory method of reducing energy demand is yet to be found, but some propositions have already been presented in the Future Work section, based on experiences with the developed prototype.
The proposed system proved capable of performing the additional functionalities intended for it. As a prototype, it still has room for refining and introducing improvements (such as incorporating an animal recognition system into it), but already in the current state of research it is compelling, that the idea of developing an efficient and highly dedicated system for automated observations in the Arctic is sound, and the goal is achievable. We hope that this paper provides a solid base for it and sets a starting point for conducting further work on more robust approaches to environmental data collection in the Arctic regions.||en_US