Open Dataset of Research Papers on A No-Code Platform for Smart Building Simulation: Integrating Artificial Intelligence for Energy and Thermal Comfort Prediction

Buildings The building sector contributes over 40% of global energy consumption and one-third of greenhouse gas emissions, creating a dual challenge of reducing carbon footprints while maintaining indoor comfort. Traditional physics-based simulation platforms, such as EnergyPlus, provide accurate predictions but are computationally heavy and inaccessible to non-experts. Recent advances in Artificial Intelligence (AI)—including AutoML for automated model selection, Long Short-Term Memory (LSTM) networks for energy forecasting, and Reinforcement Learning (RL) for adaptive HVAC control—offer high predictive accuracy but remain code-intensive and difficult to implement in practice. This study develops a no-code smart building simulator that integrates AutoML, LSTM, and RL within a drag-and-drop interface, enabling accessible modeling of energy consumption and thermal comfort. A monitoring dashboard visualizes forecasts, comfort indices (PMV, PPD), and AI-driven recommendations while embedding compliance with Indonesia’s SE PUPR 22/2024 on Smart Buildings. Experimental results show that the simulator achieves reliable forecasting accuracy (RMSE < X, MAE < Y), reduces computational time compared to baseline simulations, and maintains thermal comfort with PPD ≤ 10%. The novelty of this research lies in unifying three elements rarely combined in prior studies: (i) AI adaptability through AutoML, LSTM, and RL; (ii) no-code accessibility for non-technical stakeholders; and (iii) regulatory compliance within national smart building standards. This integrative framework bridges the gap between research-oriented AI models and real-world building management, contributing both to sustainable building operations and policy-driven implementation.