Design a large-scale temperature sensor system

System design: large-scale temperature sensor collection and visualization

Design a system to collect and display temperature data from a very large number of sensors distributed across a wide geographic area (e.g., Washington State).

Inputs / devices

• 1 million to 10 million temperature sensors, roughly uniformly distributed.
• Sensors are low-cost and simple .
• Sensors do not support mesh networking .
• Sensors send readings to the backend via HTTP (both sides have IP addresses).
• Temperature changes slowly, but the system must be able to track any sensor’s temperature over time at 10-second granularity .

Outputs / product requirements

• A near real-time temperature distribution map for a Web UI.
  • At ~1M sensors, the UI can show roughly 1 sensor per pixel on a 4K display.
• Support queries over a time range (e.g., 1 day or 1 year ) to retrieve:
  • current temperature
  • minimum temperature
  • maximum temperature
• The system should run reliably for years .

What to cover

• High-level architecture, APIs, storage, and data flow.
• Scalability, reliability, and operational concerns.
• How you serve both near-real-time visualization and long-range historical queries.