An IoT solution consists of interconnected hardware and software components that collect, transmit, and process data for decision making.
Device Layer
Physical layer. Hardware components embedded in physical devices. Low computing power, low energy consumption, lower reliability per unit. Deployed in large volumes. Handles raw data collection and direct physical-world interaction.
- Sensors
Take inputs from the environment and convert them into data. - Actuators
Modify the environment based on system commands. - Hardware interfaces
- Microcontrollers
- Embedded software
Edge Processing
Data processed at IoT devices, close to the source of data generation. Computing power is very limited. Reduces latency and bandwidth usage by processing locally.
Communication Layer
Intermediate processing layer between edge and cloud. Transmits data between devices and servers. Reduces latency by processing locally. Reduces bandwidth load on the network.
Handles:
- Local computation
- Local decision-making
- Data pre-processing
Examples:
- Wireless communication protocols
- Internet connectivity
- Networking infrastructure
Fog Processing
Data processed at intermediate nodes between edge and cloud. Nodes can be local servers, gateways, or routers. Slightly more computing power than edge devices. Reduces latency and bandwidth usage compared to cloud processing.
Cloud Layer
Back-end infrastructure (Google Cloud, AWS, Azure). Not suitable for latency-sensitive decisions. Does compute-intensive processing.
- High-power compute
- Large-scale storage
- Big data analytics
- AI/ML workloads
- Computer vision
- Voice recognition
- NLP