ETH Zurich : Computer Science : Pervasive Computing : Distributed Systems : Education : Student Projects : Abstract

A Resilient Activity Collecting System in Residential Homes (B)

Status: Abgeschlossen

Motivation

Attempts at achieving energy efficiency in households have relied on making household residents aware of their consumption. This has proven to be difficult due to the obstacle presented by the energy jargon (watts, joules, etc.) and the over- or under-estimation of the consumption of most electrical appliances in households. However, linking energy-consuming appliances to activities carried out by the residents would be more relatable due to targeting behaviors instead of actions.

Obtaining energy datasets is a tedious and costly effort, where houses need to be fully instrumented. Often, the collected data consist in power data, but higher level metadata such as when and who turned on a given appliance or what type of activity (cooking, etc.) are not available. Obtaining these quality annotations represents a challenge in terms of accuracy: how to make sure that the users annotate it religiously (beginning and end) and that the least user annoyance is caused?

Objectives

The goal of this project is to develop a fully working and resilient prototype for collecting activity data from a residential home. The work should be packaged such that an off-the-shelf deployment in another home should be possible without considerable reconfiguration of the system.

Requirements:

This project requires interest in data mining, database systems and sensor nodes. The student is expected to be familiar with relational database architectures (preferrably PostGreSQL), Unix systems and scripting (Python). The student should be interested in working with new tools and have good coding practices (coding style and documentation of code). Students are expected to be highly motivated to work on their topics and to cooperate with their supervisor regularly to discuss current progress and next steps.