Edge & Cloud Software Development
Optimizing data transmission for vehicle gateways through intelligent compression and edge computing.
Problem
Supporting the capture of as much bus data as possible on a provided, cost-effective IoT device.
Challenge
The vehicle gateway captures approx. 2 GB of bus and sensor data per hour. However, only 50 GB of upload volume per month are available for the cloud.
Edge-Based Signal Analysis
We engineered an algorithmic filtering architecture for the edge layer that evaluates the relevance of incoming sensor signals in real time. This systematic approach allows the raw data stream to be compressed to 20 MB/h while fully preserving the information density required for downstream analyses.
In-Vehicle Query Engine
Rather than streaming raw data to the cloud, we shifted the evaluation logic directly to the data source. We implemented an in-memory query engine on the vehicle gateway that executes SQL-like analyses locally. Consequently, only condensed insights are transmitted, minimizing bandwidth utilization and significantly reducing latency.