Wireless Sensor Network With Cloud Server Prototype Build and Analysis

Built a Ruby on Rails web application to receive data from a ZigBee sensor network with a cloud gateway and display in an easily understandable visual format.

The map icons corresponded to either a ZigBee Controller (C), or a ZigBee End Device (S). Map icons would populate the satellite map in real time corresponding to the most recently received GPS data from each End Device or Controller. Icons would display the most recently received sensor data in a hover box above the icon when clicked by a user, as well as in a summary table underneath the map.

The map had the ability to render multiple complete ZigBee networks differentiated by their unique Controller ID. Networks could be separated by an arbitrarily large geographical distance. The user could then highlight a particular region of interest to gain more information about a particular network.

Experience with web application development, Ruby on Rails, MVC arcitecture, SQLite databases

Designed built and tested hardware system for a battery powered ZigBee End Device reporting GPS and local temperature to a ZigBee Coordinator unit acting as an internet gateway. ZigBee star network achieved with Xbee units. ZigBee End Devices used Arduino as an embedded platform, Raspberry Pi as ZigBee Coordinator. Prototype firmware for End Device in C/C++, Coordinator firmware written in Python.

Experience with micro-controllers, embedded systems design, circuit design, sensor integration, C/C++, Python, scripting, hardware and software testing.

Ran multiple linear regression analysis to determine power consumption of ZigBee End Devices over the four major runtime modes (boot, transmitting, receiving, deep sleep). Generated estimates for power consumption within 5% error.

Experience with statistical analysis, model fitting, parameter estimation

This project was part of my M.S. Thesis, full written report available here