Watch a field-device message become a cloud-ready IoT message.
Select a gateway scenario, then play or step through the translation path. The animation separates protocol parsing,
value normalization, identity mapping, security bridging, buffering, and final delivery so learners can see exactly
where gateway value is added and where assumptions can break.
IngestCurrent translation stage
14 msTeaching latency estimate
MappedDelivery semantics status
MQTTTarget side protocol
TryChoose Semantic for Mapping profile, select Bridge TLS/DTLS on target side, and press Step through a Modbus-to-MQTT translation.
ObserveThe output retains engineering units, timestamp source, and device identity while the gateway changes register addressing into a topic schema. During a short target outage, the 30-message buffer delays forwarding instead of discarding meaning.
ExplainA protocol gateway parses the field message, normalises its semantics, maps those values into the destination schema, applies the chosen security boundary, and only then forwards or stores the result.
Technical boundariesThe trace simplifies protocol timing and assumes the configured mappings are correct. It does not model certificate lifecycle, persistent-store failure, unbounded outages, duplicate side effects, device-clock error, or physical confirmation of actuator commands.
The gateway terminates the source protocol enough to parse addresses, function codes, resources, or advertisements.
2. Preserve meaning
Raw values need units, calibration, timestamps, device identity, and a stable information model before forwarding.
3. Rebuild delivery
MQTT QoS, CoAP confirmable messages, HTTP status codes, and Modbus retries do not map perfectly.
4. Check the boundary
Security may terminate at the gateway, so trust, credentials, and audit logs must be designed deliberately.
Scenario
Translation settings
Playback
Quick controls
Field device
Modbus meter sends a register response.
Source sideModbus RTU
M2M gateway translation stack
The gateway parses the field protocol, normalizes meaning, then builds a target message.
Cloud application
MQTT broker receives a topic message.
Target sideMQTT
Frame
1. Ingest frameThe gateway receives the source-protocol message and checks basic validity.
Ready to translateMeaning is preserved when register/object/resource identity is mapped deliberately.
Payload growth8 B source -> 96 B target, because metadata and units were added.
Latency model
14 ms
Parser, mapping, security, queue, and retry settings combine into a teaching estimate.
Reliability bridge
Mapped with caveat
The gateway can retry and buffer, but source and target guarantees are not identical.
Security boundary
Gateway terminates uplink TLS
The gateway becomes a trust boundary and must protect credentials and logs.
Incoming frame
Device-side protocol format before interpretation.
Canonical object
Normalized meaning inside the gateway.
Outgoing message
Cloud-side protocol format after mapping.
Message trace
Each trace line activates as you step through the pipeline.
What changed?
The gateway changed address model, payload format, and delivery contract while preserving the temperature meaning.
Learner checkpoint
Ask: what would break if the register scaling factor or topic mapping were wrong?
Learning Support
Use these cards after the animation to connect the moving message to design decisions.
Translation is layered
A gateway may bridge physical media, sessions, addresses, payloads, and application meaning. This animation focuses on the application-level part where most mistakes happen.
Semantics matter
Register 40017 or a BLE UUID is not enough for cloud analytics. The gateway must attach units, scaling, identity, and context.
Reliability is not copied
MQTT QoS, CoAP confirmable messages, HTTP retries, and field-bus polling are different mechanisms. The gateway must document the effective behavior.
Security shifts trust
If the gateway decrypts or signs on behalf of devices, it becomes a high-value component that needs hardening, audit, and secret protection.
Quick Reference
Common source side
Modbus: registers, coils, function codes, polling
BACnet: objects, properties, building automation events
HTTP/REST: endpoints, status codes, request retries
CoAP: constrained REST, UDP, optional reliability
OPC UA or platform APIs for richer information models
Gateway checks
Is device identity stable across protocols?
Are units, byte order, and scaling correct?
Can duplicate messages be detected safely?
What happens during cloud outage?
Design warning
Do not describe a protocol gateway as a simple converter unless semantic mapping, reliability behavior, and security boundaries are intentionally out of scope.
Technical Accuracy Notes
The latency and byte values in this animation are teaching estimates, not product benchmarks. Real values depend on
gateway hardware, source polling rate, serialization format, broker/cloud distance, TLS session reuse, queue policy,
and target service behavior.
Modbus register values require the correct function code, byte or word order, scaling factor, and device register map.
CoAP and HTTP can map conceptually because CoAP uses a REST-like request/response model, but retransmission and caching behavior still differ.
MQTT QoS is broker/session delivery behavior. It is not the same as confirming that a physical actuator completed an action.
oneM2M-style architectures treat the service layer separately from underlying network transport. Gateways often live at that boundary.
Practice Prompts
Try these before moving on.
Find the semantic step
Run the Modbus scenario with the raw tunnel profile. Which cloud-side information is now missing or ambiguous?
Test outage behavior
Set the buffer to 0 and retries to 0. What would happen if the cloud endpoint is unavailable for one minute?
Control path check
Use the MQTT-to-CoAP command scenario. Why does a delivered command message not automatically prove the actuator moved?
Related Learning
Use these pages to connect the animation to wider gateway and protocol topics.