In Saudi Arabia, AI-guided building controls cut electricity use in desert heat

An 18-month field study found lower energy use in eight occupied buildings, including during extreme heat and dust storms.

Editorial illustration — not from the study.

Researchers tested an adaptive control system in six residential units and two small commercial buildings over 18 months. The system used a continuously updated digital model of each building and short-term predictions to adjust heating, ventilation and air-conditioning settings in response to changing conditions.

What the controls achieved

Compared with conventional thermostat operation, the system reduced electricity use by 37.8% while maintaining the study’s ASHRAE comfort range during 93.5% of occupied hours. Savings were 36.5% in the residential buildings and 41.8% in the commercial buildings. The evaluation included outdoor temperatures above 45 °C and 48 recorded dust-storm events. During dust events, anticipatory control increased savings to 44.2% in residential buildings and 46.8% in commercial buildings. The system also outperformed model-predictive control and a local reinforcement-learning system without coordination across buildings. Transfer learning reduced commercial training time by 38% and computational demand by 56%. Separate simulations based on the calibrated building model projected 41.2% energy savings and a 43.8% peak-load reduction across 150 buildings; these projections were not confirmed by an additional field deployment.

Why desert buildings matter

Cooling buildings in hot, dusty desert cities can require large amounts of electricity, while dust storms can affect short-term indoor conditions and HVAC performance. The field results suggest that climate-adaptive controls can reduce energy use while maintaining the specified comfort range and responding to dust events. The findings are directly relevant to cooling-dominated buildings, but broader testing is needed before the results can be applied confidently to larger or more varied building portfolios.

Evidence and remaining limits

This was an 18-month field evaluation in eight occupied buildings in Hail, Saudi Arabia, supplemented by simulations for a 150-building community. The field comparison used conventional thermostat operation and also included comparisons with other automated control approaches. The study took place in one hot-arid city and involved six residential units and two small commercial facilities, so its results may not represent other climates, building types or operating conditions. The projected community-scale savings came from simulation rather than an additional field deployment. The researchers also state that validation in larger and more diverse building portfolios is still required.

An AI design tool helped teams spot more possible harms early

A study of 18 cross-functional practitioners found that an AI-assisted design tool helped teams identify more potential harms than standard worksheets. The tool used shared planning blocks, checklists and simulated user perspectives to support early discussion of responsible AI concerns.

2 days ago·14 views

Engineering & Technology

Ammonia could cut coal-power emissions, but nitrogen oxides remain a hurdle

A review of laboratory, industrial and computer studies finds that co-firing 20%–30% ammonia with coal is feasible in existing coal-fired boilers. Higher ammonia shares may raise nitrogen-oxide emissions and lower boiler efficiency, making fuel-injection design and air staging important.

3 days ago·14 views

Engineering & Technology

Using satellite-based altitude in cruise could burn more fuel, study finds

A simulation study found that using geometric altitude instead of barometric altitude during cruise flight would, on average, increase fuel use. It also found a considerable risk of exceeding aircrafts’ maximum barometric altitude, potentially requiring more step descents.

4 days ago·18 views