Original title: Experimental Demonstration of Cost-Effective Real-Time Risk Prediction and Energy Management System for Fog-Cloud-IoT Architecture in Smart City Applications
Sami Abduljabbar Rashid, Yousif Ismail Mohammed Al Mashhadany, Ali Amer Ahmed Alrawi · September 2026
Smart cities rely on connected smart homes, but sending all device data to distant cloud servers causes high delays, uses too much power, and slows down threat detection. To solve this, researchers designed a system named CREFCI. It places a local computer, called a Raspberry Pi 4 fog nodeA small computer placed close to the devices that handles processing locally instead of sending everything to a distant cloud., directly inside the home to process data nearby instead of sending everything away. The setup translates signals between Wi-Fi, Zigbee, and Z-Wave devices, checks for network threats with security software, and uses artificial neural networks to predict power needs. It also syncs with the cloud so users can check apps. The tests showed major gains. CREFCI lowered energy use by 50–80%, consuming just 0.3–0.5 kWh instead of 1.0–1.7 kWh. It reduced data lag by 40–60%, reached 75% accuracy in energy predictions, and cut average prediction errors by 50%.
Imagine cooking dinner for guests. You normally check the recipe once, then run to the store for every missing ingredient, and the whole meal waits on each trip. Now suppose the guests' needs kept changing while you cooked, so a plan made in advance was already out of date. In this paper, a smart home faces the same problem: energy use and security risks change quickly, so waiting to send every decision to a distant central office is too slow. It keeps a small helper nearby to make quick local decisions and only calls the main planner when needed.
For smart homes and cities, this means lower electricity bills, faster responses, and fewer security gaps, because data is processed nearby instead of only in the cloud. It also makes real-time risk prediction practical, so threats can be caught while they are still happening. That matters for anyone relying on connected devices to be both affordable and safe.
The abstract doesn't say how many homes or devices were tested, which existing models were compared, or how the system handles rare threats. It also reports 75% accuracy, so a quarter of energy predictions are wrong, and cost-effectiveness is claimed without a detailed cost breakdown.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
Experimental Demonstration of Cost-Effective Real-Time Risk Prediction and Energy Management System for Fog-Cloud-IoT Architecture in Smart City Applications — Sami Abduljabbar Rashid, Yousif Ismail Mohammed Al Mashhadany, Ali Amer Ahmed Alrawi — F1000Research — doi:10.12688/f1000research.176542.2
Abstract licence: CC BY
Explanation generated by Synthesis from the abstract — it is not the original paper.