Project Group Number 4: SmartMON

Group Members: Hieu Nguyen, Teija "Aaron" Zhang, and George Mansour, Ryan Bald

SmartMON's Website: http://cpmqtt1.calit2.uci.edu:3000/ (might be under construction)

SmartMON: A Device Usage & Tracking System for Ed-Tech Admins

Introduction:  Education spending for K-12 is at historical highs. As seen from the figure below, educational spending has averaged about 40 billion dollars over the past 10 to 15 years. One of the largest contributors to this spending is on IT. In 2015 U.S education institutions spent 6.6 billion dollars on IT. The use of technology is extremely common in education and various other fields. The majority of schools in the U.S have technology such as computers, TV's, projectors laptops, and tablets in the classroom with an aim to enhance the student's ability to learn. Unfortunately, many of these expensive resources often go without full intended usage.   Expensive training for integration of technologies  educators may not be implemented.  The concern could be a mismatch in expectations between Education spending for K-12 is at historical highs. As seen from the figure below, educational spending has averaged about 40 billion dollars over the past 10 to 15 years. One of the largest contributors to this spending is on IT. In 2015 U.S education institutions spent 6.6 billion dollars on IT. The use of technology is extremely common in education and various other fields. The majority of schools in the U.S have technology such as computers, TV's, projectors laptops, and tablets in the classroom with an aim to enhance the student's ability to learn. Unfortunately, many of these expensive resources often go without full intended usage.   Expensive training for integration of technologies  educators may not be implemented.  The concern could be a mismatch in expectations between administrators, educators, and technical staff.  Verbal and written assessments from educators may not fully provide the depth of the problem.  Depending on the site or district political climate, a problem may not even be shared to administrators or technology administrators. The success of pilot deployments may not be able to be fully assessed ahead of scheduled rollouts.  Could there be a way to independently audit technology usage to make smarter education tech. decisions?  A simple solution to this problem is what this project intends to provide.  

Goal Statement:  This project seeks to develop a low cost auditing system to provide visibility into the utilization of technology purchases for education.  By monitoring power consumption, device usage can be inferred, analyzed, and displayed to provide useful information about the adoption of educational training programs or the usage of new devices.  Problems in deployment can be caught and corrected earlier in pilot sites  before mass deployment occurs.  This system would not only provide a real time dashboard showing recent device usage and trends, but would also automatically author usage reports.   Analysis of usage trends automatically occurs within the device server-based analytics engine. The information provided by this system can augment information gathered in discussions between administrators, technology staff, and educators to provide full visibility for the effectiveness of technology implementation. 

The system is comprised of two parts:  1) A low cost hardware module that plugs in-between the device and the wall, and a server based analytics engine.   Multiple devices may be plugged into a single hardware monitoring module.  Many modules can be deployed simultaneously across multiple sites.  2) A cloud based server analytics system that gathers, stores, and analyzes data.  This system has a web-based interface to clearly provide the output of analytic analysis to the  end user to assist in actionable technology decisions.

Hardware:  Our main goal for input in the first step of the prototype was to retrieve simple current  (and eventually RMS Power)from a monitor device. We chose to use the Particle Photon™ as our microcontroller unit as it is able to provide us with the WIFI capabilities we need, it also works hand-in-hand with a demonstration current sensor module from ControlEverything. 

Servers:  To host our front-end and back-end software programs and MQTT (communications system) broker, we use two Ubuntu servers. Server number one acts as the sole provider for MQTT. We installed Mosquitto (a private/local MQTT broker) on both servers but number one is used for all communications between the Photon and Node.js. The second server is used as our main development server. We installed MEAN Stack on it and created both our front-end and back-end on this server.

Software MEAN Stack:  Our application will run on top of a MEAN stack (MongoDB, Express, Angular, Node). MongoDB is the database we are using to store all of our pertinent data (mainly the amount of current running through a wire). Node.js will be used as the server side to our application (serving requests to our application’s webpage, etc.). Angular.js and Express.js will be our front-end and back-end frameworks, respectively.

Current Progress: The current status of our Smartmon device is in the later analytical development stages. Currently the device is able to monitor the real time rms current, voltage, power factor, current summation, and instantaneous demand. From these measurements, we have derived various algorithms to analyze the data and have outputted the analytical graphs onto a webpage. So far, for our analysis, we have been successfully able to output the average of Smartenit values for the past 12 hours, the total on time vs off-time per work day, the total time on per day vs week day, the times turned on in a weekday, percentage of the most used operation mode, and all working conditions. These analytics are in real time but are currently only for the measurements of a PC. Our users will be able to plug their PC into the Smartmon device and will then be able to see various graphs that analyze their overall usage of that device. Our next task is to come up with the analytics for devices such as a projector and a T.V. The challenge here will be coming up with the best graphs to analyze the data most efficiently as they pertain to these devices.