Architecture Layer Builder Game
Sort IoT components into device, network, processing, and application layers
Architecture Layer Builder Game
Build an IoT architecture by placing each component where it mainly works. The same data moves from physical devices, through connectivity, into processing, and finally into user-facing applications.
Architecture layer building controls and outputs
Level 1: Smart room climate monitor
Start with obvious parts: a sensor reads the room, network pieces move the reading, cloud services store and evaluate it, and applications show the result.
Available components
Select one card, then choose the layer where that component mainly belongs.
Architecture canvas
Place the selected component into a layer. Correct cards stay in place; wrong cards stay available so you can try again.
Try the temperature sensor first. It belongs in the layer that measures or acts on the physical world.
Smart room climate monitor
A room node measures temperature, sends readings through Wi-Fi, stores time-series data, and alerts a facilities user when comfort drifts.
- Data starts as: Raw temperature reading
- Data becomes: Dashboard trend and comfort alert
- Design goal: Keep each responsibility in the right layer.
Selected card
No component selected yet.
- Clue: A device-layer part touches the physical world.
- Why it matters: Layering makes systems easier to scale, troubleshoot, and replace.
What to notice
- Physical interaction belongs at the device layer.
- Data movement and protocol translation belong at the network layer.
- Storage and analytics belong at the processing layer.
- User decisions and business workflows belong at the application layer.
Plain-language model
Ask what the component mainly does. If it measures or moves something physical, it is device. If it moves data between places, it is network. If it stores or transforms data, it is processing. If a person or business process uses it, it is application.
Technical accuracy checks
- A component can be implemented in different places, but this game classifies its primary responsibility.
- An MQTT broker may run in cloud infrastructure, but in this simplified four-layer model its role is messaging and connectivity.
- Edge processing can sit near the network/gateway boundary; when the main role is filtering or inference, classify it as processing.
- Dashboards do not become processing just because they show processed data.
Try these checks
- Complete Level 1 without using Hint.
- Use Level 2 to separate gateways from cloud services.
- Use Level 3 to spot the difference between ML inference and an operator console.
Common mistakes
- Putting dashboards in processing: dashboards display decisions; processing creates the data products they display.
- Putting gateways in device: gateways may be hardware, but their architectural job is connectivity and translation.
- Putting ML models in application: the model produces insight; the application presents or acts on that insight.
Layer quick reference
- Device: sense, actuate, power, collect raw signals.
- Network: connect, route, bridge, translate protocols, broker messages.
- Processing: store, filter, aggregate, infer, normalize, manage device state.
- Application: visualize, notify, report, trigger workflow, support user decisions.
Independent prompt
Pick one component you placed correctly. Explain what data enters it, what data leaves it, and why that responsibility belongs in that layer rather than the layer above or below.