Advisor: Jeff Ondich
During the 2017-2018, 2018-2019, and 2020-2021 school years, CS comps teams worked on the topic of Energy Analytics. Here's an excerpt from the descriptions of those earlier projects.
Carleton's energy facilities heat and cool the buildings, turn on the lights, provide hot showers, and power the labs. These same facilities also generate enormous amounts of data. What temperature is it in Weitz 225C? Is the fan on above the CMC 3rd-floor office hallway? How much power is being generated by the wind turbines? How much power are we drawing from the grid? Which of the hundreds of outside air dampers are open or shut and how much?
In 2011, Carleton committed publicly to its Climate Action Plan, whose central goal is carbon neutrality for the campus by 2050, with intermediate emissions goals set for 2020, 2025, and 2030. As part of our pursuit of these goals, Carleton has also created a Utility Master Plan, which involves, among other things, geothermal heat pumps, solar power, and conversion of the campus heating systems from steam to lower-temperature water. One of the first and biggest steps of the Utilities Master Plan will be the installation of a new energy station beneath the new science building after Mudd is demolished this coming fall. The Utilities Master Plan is a long-term investment intended to support responsible management of the campus for a century or more. [April 2020: Mudd is long gone now, replaced by Anderson Hall and the energy station beneath it.]
Making good on our commitments to the Climate Action Plan is a complicated business, involving human challenges as well as technical challenges. Martha Larson, Carleton's Manager of Campus Energy and Sustainability, works with those challenges every day, and she would like to add some tools to her toolbox.
One such tool is the relatively new field known as Energy Analytics, which concerns itself with using data analysis to reduce the energy costs of running buildings and groups of buildings.
For this project, you will apply energy analytics techniques to Carleton's energy data in hopes of generating insight into how we could more efficiently use our energy resources.
Well-applied Energy Analytics techniques could help the Facilities Department to automate the detection of all sorts of problems: open windows in the winter; open windows in the summer when it's 90; simultaneous heating and cooling of the same building; erratic and atypical behavior of a thermostat or a vent; broken radiators; etc. Pursuing that kind of automation is the goal of this comps project.
When the 2020-2021 team concludes their work in a few weeks, here's what we will have:
This project will build on the work of the previous three comps teams with the goal of creating a tool that the people in the Facilities department would actually choose to use to help them manage Carleton's buildings. That's an ambitious goal, since professionals aren't going to use a tool unless it helps them do their jobs better or more easily.
With that in mind, this project will focus on:
There are no mandatory prerequisites for this project. That said, it would be great if one or two members of the team included some expertise in statistics (STAT 120), artificial intelligence (CS 321), or machine learning (CS 320).
Dozens of companies that deal with energy analytics have online introductions to the topic. These introductions tend to be vague, and mainly concerned with convincing people to buy their products. Typical examples include this one from Siemens, the company whose tools manage energy in about half of Carleton's buildings, and this one from IBM.