Final Exam - Spring 2009

Its all about me ...


In the past I've had students collect their own data for the final exam and then conduct investigations to answer some questions that I had, like "does the refrigerator get warmer inside if you leave the door open?" or "where are the rogue wireless networks in ERF?" or "what is the magnetic field around a microwave oven?"

This year I found myself asking questions about electricity / water / natural gas usage at home. How much am I using? Have any of the improvements I've put into place (CF lightbulbs, programmable thermostat, etc) made any obvious difference or have they been outweighed by buying a bigger TV and putting in a pond? At the same time I was also curious to know what you could learn about a person from just their utility bills. I decided to combine the 'green' topic and the 'cyber-stalking 101' topic together into this final exam.


I've kept all my old utility (electricity, natural gas, water) bills in a file cabinet in the basement, but I've never actually looked at the numbers, so I grabbed them and typed in the relevant information to create a table containing 10 years of data on:

The data is available here: http://www.evl.uic.edu/aej/526/526finaldata.csv

Electricity is typically billed once a month, but typically every other meter reading is an estimate. Water is typically billed every 2 to 3 months, and natural gas every 1 to 2 months. This makes it harder to see short term trends, but longer term trends should still be visible.

The temperature data in the file comes from the electricity company. I'm not sure how much I trust that data. You might want to get temperature data from another source, and probably in addition to the average also look at the minimum and maximum.


You should start by looking at data itself and do some simple plots in your favourite spreadsheet / plotting program. There will be some obvious cyclical patterns such as air conditioning driving up electricity usage dramatically in the summer and the furnace driving up natural gas usage dramatically in the winter.

Your job is to look beyond the cyclical patterns for longer term trends and aberrations that are hiding there, and see what changes in the real world could have caused them. Some of these changes are related to hot hot a summer was or how cold a winter was, others are related to human behavior. This is where the cyber-stalking / privacy issues part of the project comes in. If you have access to this utility data and can filter out the repeating patterns, and the general environmental changes, can you find interesting events or trends that tell you about the people?

The goal here is to create a set of effective visualizations to aid in your analysis and to back up any conclusions you draw.



- The house stayed basically the same throughout the 10 years of data collection.
- The house uses electricity to run the air conditioner for cooling, and to run the blower on the furnace for heating
- The house uses gas for heating and cooking, and drying laundry


I did some measurements of my electricity usage last month in the spring:

In a common house:
This data should allow you to break up the daily usage into components that can vary over the months and years.



Here are some relevant facts that should give you some things to look for. Many of these I don't remember the dates for. Maybe you can come up with some guesses as to when they happened.

- We replaced our top loading clothes washing machine with a front loading machine
- We switched the house over to CF bulbs
- I had an SGI indy that was running all the time that I turned off. It was replaced by a PC that ran all the time, which was turned off a couple years later. Now all the machines are mac minis or laptops which take much less power.
- There is only one TV in the house but it was replaced/upgraded twice during those 10 years.
- We started using a programmable thermostat to make the air-conditioning and heating less aggressive when no-one was expected to be home
- We started gardening aggressively which used more water. Later we started using mulch and began using much less water for the plants
- We started a 400 gallon pond in 2000 and expanded it to 650 gallons in 2002. Ponds require complete water changes at least a couple times a year. In 2002 we did many water changes as we had a bunch of sick fish in the pond - this is really obvious in the water data.
- We put in a new (more efficient) air conditioner
- We had a couple people staying with us for 3 weeks.

You will very likely need to look up some other reference material on the web about water, gas, and electricity usage. I did. I also found that some sites over-inflate the usage numbers by 50 to 100% compared to the numbers that I was able to measure. I would recommend looking at a few different sites to get a better feeling about what good average numbers might be. Be sure to cite the websites that you use.

You can use whatever software you want to extract the patterns in the data and to visualize the results. Please document this work in your web page. The web page should show visualizations of the data itself, and visualizations of divergence from the norms. Show all the valuable visualizations that you created and used to do your analysis work.

Task #1 - document the repeating seasonal patterns and what affects those patterns (e.g. temperature)

Task #2 - given the results from task #1, document the long term trends and short term variations in those patterns (e.g. given temperature readings and a spreadsheet program, you should be able to work out expected values for gas and electricity usage, and then any variation you see might suggest a human cause.)

Task #3 - given the results from task #2, try to make some hypothesis as to what human activities caused those long term trends or short term variations.


This exam is going out during the Project 3 presentations, and during that time people will have the ability to ask questions in class about the data to try and clarify any questions you have that would help your analysis. I think I've given you all the relevant facts, but maybe I've forgotten something. As you look at the data maybe you will find something in a particualr month/year - ask me about it and I'll see if I can go back through my records and find some event that may be applicable.

Here are some of the dates that I have found:
June 2000 - Air conditioner replaced
July 2004 - New programmable thermostat installed
July 2006 - 2 week trip out of the country
Fall 2006 - Started using mulch and less water in the garden
2007 - Transition to compact fluorescent bulbs
March 2008 - Front loading washer repalces top loading washer



By the beginning fo the final exam period each person should create a web page of their results. Again as usual everyone should send me a 320 x 240 jpg snapshot for the web.

There is no software to turn in for this project, just a series of screen snapshots (possibly from different pieces of software)

During the final exam period itself each student will have 10 minutes to describe their visualization work and their findings.

last revision 5/8/09