Abstract
This article examines concepts of human–machine interaction that support a human-centered approach: cybersocial systems of Industry 5.0 and symbiotic systems. It analyzes the potential for creating symbiotic smart home systems. It also looks into the most common tasks solved by smart homes in the area of electricity consumption, such as analyzing and forecasting energy consumption within the system and developing recommendations for smart home ecosystem participants based on this analysis and forecast. The mutual influence of these tasks in a symbiotic smart home system is demonstrated. It proposes an approach to identifying typical energy consumption scenarios and interpreting consumer behavior. It develops an ensemble learning algorithm for time series forecasting based on clustering. The article also explores the feasibility of implementing the concept of rational consumption and generating recommendations for shifting consumption to periods with lower tariffs based on household electricity consumption forecasting.