<p>  </p> <p>Heating and cooling (HC) energy use account for about 40% of the total annual energy consumption and cost of an average household in the U.S and it is significantly affected by residents’ energy-related behavior. This is particularly important for low-income residents in the U.S. who spend a larger portion of their income (i.e., about 16%) on home energy costs compared to average-income households (i.e., 4%). To address opportunities for reducing residential HC energy usage without requiring physical building upgrades, this thesis presents a new paradigm for smart and connected energy-aware communities that leverage smart eco-feedback devices and social games to engage residents in understanding and reducing their home energy use.</p> <p><br></p> <p>First, this Thesis presents a new modeling approach for personalized eco-feedback design integrated with a collaborative social game to assist residents to enhance their thermostat use while promoting community-level energy savings. The modeling framework is integrated into a cloud-based application, MySmartE, with visual (wall-mounted tablet) and voice (Alexa) user interfaces to facilitate behavioral changes in a user-centric approach. The platform is deployed in a multi-unit residential community in Fort Wayne, IN and the experimental data are used to investigate: (i) how occupants’ thermostat behaviors changed after using the MySmartE app; (ii) how users interacted with the app during the game; and (iii) how was users’ experience with the developed platform. Despite the heterogeneous characteristics of households, the results from the field study show the positive effect of the intervention in the thermostat-adjustment behaviors, which results in an increase in the indoor temperature during the cooling season compared to the baseline period. Findings from the user interaction analysis and post-experiment interviews also reveal the significant potential to nudge households’ energy conservation behaviors with the developed platform along with the challenges that should be tackled to derive long-term behavior changes. </p> <p><br></p> <p>Second, this Thesis introduces a sociotechnical modeling approach based on utility theory to reveal causal effects in human decision-making and infer attributes affecting households’ thermostat responses during an eco-feedback intervention. This modeling approach (i) is based on a utility model that quantifies residents’ preferences over indoor temperatures given decision attributes related to their thermal environment and eco-feedback and (ii) incorporates latent parameters that are inferred to determine the unique behavioral characteristics of each household. For parameter learning, a hierarchical Bayesian model is developed with a non-centered parameterization and calibrated to the field data. Based on the calibration results, the proposed model quantifies the impact of the eco-feedback on households’ thermostat-adjustment behaviors and serves as a foundation for analyzing resident behavior in connected residential communities with eco-feedback energy-saving programs.</p> <p><br></p> <p>Finally, this Thesis presents a modeling approach for investigating the decision trends of residents in goal-oriented collaborative social games while considering their decision preferences and goal achievement capabilities. The proposed approach involves a mechanism design method that derives optimal decisions by conducting counterfactual simulations given various scenarios of goal and reward sets. This modeling approach (i) re-defines utility functions to include decision attributes that reflect user preferences on the game status; (ii) calibrates the model to learn the decision preferences of the residents; (iii) simulates the decision-making process of residents by solving the Nash Equilibrium for a given set of game scenarios. The results revealed the decision trends of the residents given the various goals and rewards along with the potential goal achievement trends and the resulting variations in the marginal community utility.</p>