AI Nudges and the Future of the Energy Grid

Abstract

With recent increases in smart grid adoption, high-frequency data for individual devices represents a wealth of opportunities for grid operators to optimize both the supply and demand of electricity. In Exploratory Analysis of Energy Patterns from Smart Grids, the first study examines such datasets using dynamic time warping, a technique that has not been used in the literature thus far. Results indicate devices are used differently across geographical locations and suggest grid operators should focus on reaching out to outliers. the second study proposes a Nudging-Based Taxonomy for Energy Consumption and incorporates nudge theory into the present, future, and distant future of electricity management, with a focus on home energy management systems that integrate various smart devices into a single system. These systems can build detailed profiles of consumers and tailor nudges to users, allowing for more efficient electricity use. The third study focuses on Understanding the Role of Trust and Personal Norms for AI-Generated Mobile Alerts. We examine the behavioral intent towards AI-generated nudges through protection motivation theory via appeals to environmental harm. Personal norms and AI trust belief were found to motivate to adoption of the proposed system, while fear was not found to be significant, suggesting the cognitive path is far more important than the emotional path.