The integration of Electric Vehicles (EVs) with renewable energy, particularly solar power, is crucial for reducing fossil fuel reliance and optimizing residential energy use. This chapter introduces a novel approach to scheduling EV charging and discharging in a solar-equipped residential building. Using one month of historical data, day-ahead solar generation and grid supply trends were predicted with a Stacked Long Short-Term Memory (LSTM) neural network. A rule-based heuristic optimization method then aligned EV scheduling with the Time of Use (ToU) tariff structure. The study utilized the Ampere Reo electric scooter with a 48 V, 24 Ah lithium-ion battery, prioritizing solar energy for daytime charging and discharging during high grid load periods. This strategy maximized renewable energy usage, reduced electricity costs, and minimized grid dependency during peak hours. Results highlight the effectiveness of combining machine learning forecasts with heuristic optimization for energy-efficient management. The findings offer scalable insights for enhancing renewable energy adoption and EV integration in residential systems, advancing sustainable energy practices.