2026-06-01 SUSTAINABLE ENERGY TECHNOLOGIES AND ASSESSMENTS 2026 90(卷), null(期), (null页)
Energy-intensive cooling is required to maintain thermal comfort in hot, dry climates, but existing systems lack real-time feedback and adaptability. This study introduces a bio-adaptive system that integrates occupant biofeedback, neuromorphic computing, and energy harvesting for sustainable building operations in arid climates. It comprises a hybrid spiking neural network-transformer forecasting model, electroencephalographybased neural satisfaction index, an Intel Loihi-based spatiotemporal controller, and triboelectric nanogenerators tailored for low-humidity conditions. The system was tested for a year in three office buildings in Saudi Arabia, achieving energy savings of 40.3% to 42.2%. The forecasting model had mean absolute errors of 0.9 kW (one-hour) and 2.8 kW (four-hour). Comfort levels increased by 27.4% on average, and 35% in conference rooms. Peak HVAC load reduced by 31.4%, harvesters produced 17.1 mW, and 87% autonomy was achieved during working hours. The economic analysis reveals a 7.3-year payback period (4.9 years with projected tariff rises) and an internal return rate of 11.2%. This research suggests opportunities for integrating cognitive systems into energy-constrained buildings.