Neuro Wireless Sensor Network architecture: Cool stores dynamic thermal mapping

This work focuses on developing cool store´s thermal mapping system based on the neuro Wireless Sensor Network (nWSN). The network intelligence is taken care of by the sensor network embedded neural net. The target application of the architecture development is for cool stores with emphasis on meat storage. The meat quality is a significant characteristic within the cold chain management. Temperature is the main parameter that needs to be monitored. nWSN architecture deals with the distributed neural network (NN) that predicts the temperature variations within the space. WSN nodes have categorized into infrastructural sensor nodes and portable sensor nodes. The neural net algorithm is embedded into the fully functional infrastructural nodes while portable nodes provide the surrounding information. The portable nodes that are attached to the meat carcass are dynamically establishing connection with one of the infrastructural nodes. The overall network formulates clusters based on Dijkstra´s algorithm. The Nodes Message interaction (NMi) model is developed to organize the communication between the infrastructural nodes and portable nodes. This work disclosed the viability of nWSN architecture to execute further on a real-time test bed. It is found that the Mean Absolute Error (MAE) at the infrastructural nodes has a variation of 1°C. The resulting MAE is considerably good where nWSN can be capable of yielding similar applications of predictions.