In 2005, Americans consumed 100 quadrillion BTUs of energy, or almost six times the world wide average per person. Much of this energy consumption is directly related to personal, individual activities such as lighting, heating and cooling. By making simple modifications to these routine activities individuals can conserve energy and help reduce total CO2 emissions. This project leverages Internet technologies to develop a scalable approach to changing behavior regarding environmental footprint (the amount of natural resources or emissions required to support an individual or household). The intent is to use online social networks, cell phones and other technologies to provide a set of ecologically sound, context-sensitive behavioral recommendations. The ultimate goal is to achieve Internet-scale energy behavior change via behavioral research on successful ways to motivate change, environmental impact analysis, and ubiquitous computing that supports objective measurement and context-sensitive communication.
The project will use the existing StepGreen.org website as a testbed for three core activities. First, controlled field studies will be conducted to examine the value of personalization and social influence on motivation to change as well as how motivational strategies can accommodate differences among ethnic and cultural groups. Second, inexpensive sensing techniques will be developed that can unobtrusively extract information about changes in daily behaviors. For example, this will be able to track home heating, cooling and transportation behavior through simple financial monitoring of transactions such as electricity bills and gasoline purchases. Finally, the StepGreen deployment, as a large-scale social intervention, must adapt flexibly to contextual information about people (e.g., favored communication modalities, appropriate timing).
The multidisciplinary nature of the proposed work enables innovative changes in several scientific domains. By bringing together technical innovation and social science, the proposed work can test theories of social computing in the field, providing a detailed understanding of how online social networks in combination with behavioral intervention strategies can lead to widespread behavioral change. Simultaneously, this work will explore and innovate solutions for large-scale distributed evaluation of Ubicomp technology. By bringing together environmental science and machine learning, the investigators will create the first objective calculator of individual energy use. Finally, the research will benefit environmental decision making by identifying misconceptions about relationships between behavior and energy consumption, educating a broad populace on individual actions they can take to reduce their consumption, and, ideally, achieving reductions in U.S. energy use.
Broader impacts. The proposed research will benefit education through the interdisciplinary training of graduate and undergraduate students and through high school outreach as part of Carnegie-Mellon's Green Design Apprenticeship. The work?s focus on different socioeconomic groups will help to highlight the importance of addressing cross-cultural issues in design and behavioral science. Finally, the project will benefit society by helping to address a critical social issue and by providing results that inform other domains in which societal benefits depend on individual behavior change (e.g., health, littering, community action).
The goal of the Stepgreen project is to leverage Internet scale technologies to create opportunities for reduced energy consumption. The original vision of the project was to leverage existing online social networks to encourage individual change. Since then the project has broadened to include a number of other ideas. We have explored the impact of demographics on energy use practices; studied the value of empathetic figures such as a polar bear for motivation and explored organizational-level planning. We have also developed mobile technologies that can provide feedback about green actions on the go. Our main product is the stepgreen.org website and associated API. It provides a mechanism for allowing individuals to report on and track their environmental impact. It includes a visualization that can be displayed on an individual's social networking web page. You can try it out at stepgreen.org. We have also launched a site in conjunction with Cornell’s CALSGreen effort in October 2010 (cornell.stepgreen.org); launched a site with Zoo Pittsburgh in June 2011 (pittsburghzoo.stepgreen.org), and recently began supporting the Community College of Allegheny County. Using Stepgreen's API we have also created alternative interfaces such as a community monitor application that we deployed in 15 households along with a home energy monitor. Households in the same community could monitor each other’s average daily consumption and share knowledge and information. Households could also view detailed information about their energy use and see information about strategies to reduce home energy use. We used our deployment to understand how data sharing between households affects them. Our intervention design leverages existing behavior theory to help explain the effectiveness of our application, and build upon studies of home energy consumption by exploring social factors such as community engagement and comparisons. Sustainable behaviors are influenced by factors such as intentions, habits and contextual, or external factors. Our findings identify how these factors shape and constrain energy consumption behaviors. Our research results demonstrate: 1) how participants integrated our application into their existing routines and habits and how this led to a positive impact on sustainable behavior; 2) how households from one community identified and addressed energy-related issues discovered as a result of using the technology; and 3) how trust plays a key role in stakeholder communication and environmental behavior. Our findings were consistent with past work in that real-time feedback and information influence attitudes to perform energy conservative behaviors. Social factors, such as comparison, affected attitudes. Comparison in our study did not encourage competition, unlike in prior studies. Instead, social comparisons conveyed new norms and knowledge transfer, and as a result, influenced attitudes. We found benefits in our mobile tablet deployment as householders shared the device with each other, and some participants used the tablet on a routine basis. This routine use allowed participants to indirectly monitor their consumption due to the presence of the polar bear display. Though our deployment was limited to renters, many in low-income environments, we expect most of our findings to generalize to renters and homeowners across socio-economic statuses. However, based on our results, factors such as trust, community cohesion, presence of landlords, access to technology, and community cohesion vary based on specific community characteristics. For example, very trusting communities, home owners, and internet users may not have reacted in the same way to our intervention. Income is also likely to indirectly affect generalization for example because of its impact on technology access. In the rest of this section, we provide design implications for future community-focused home energy technologies. Specifically, we applications should leverage social factors to encourage householder engagement and community trust, and support privacy, social roles, and effective comparisons.
