Case studies / Cloud
Cloud · Maritime / logistics technology
Re-platforming an IoT vessel-tracking system to AWS — and tripling its capacity
A vessel-tracking IoT platform had hit a throughput ceiling on its original infrastructure, risking data loss and outages as the fleet using it grew. We re-platformed it onto AWS and tripled its capacity.
3× the platform's original throughput.
Held steady through the migration and beyond.
Tripled without a full product rebuild.
The challenge
A throughput ceiling the growing fleet was about to hit
The vessel-tracking IoT platform had hit a throughput ceiling on its original infrastructure. As the fleet using it grew, the risk of data loss and outages grew with it.
Before
- Throughput ceiling on original infrastructure
- Growing risk of data loss under load
- Limited redundancy for an operationally critical system
- Scaling required manual intervention
After
- Re-platformed onto AWS for elastic scale
- ~1.5 billion messages/day, 3× original throughput
- 99.9% availability maintained
- Redundancy built into the new architecture
How we approached it
Redesign for elastic scale and redundancy
We rewrote the device-to-server data path on Kinesis, ECS and Glue, moved to a serverless architecture to remove the single points of failure in the original design, and implemented MQTT with mTLS device enrollment through AWS IoT Core — while encrypting data both at rest and in motion.
Under the hood
Next step
An operational system hitting a scale ceiling it wasn't built for?
We re-platform systems for elastic scale and redundancy — without a full product rebuild.