The Challenge
What MaxSense Was Facing
MaxSense deployed environmental sensors across industrial facilities using hardware from multiple vendors — each transmitting data over different protocols (MQTT, CoAP, proprietary serial) at different sampling rates and in different payload formats. Sensor data flowed into a fragmented set of monitoring tools with no common data model. Alerting was configured separately per vendor tool, meaning cross-sensor correlation — identifying conditions that only became significant when multiple sensor types agreed — was impossible without manual analysis.
The Solution
What We Built
We built a unified IoT ingestion pipeline with per-protocol adapters at the edge gateway layer, normalising all sensor payloads into a common time-series data model before forwarding to a centralised stream processing engine. The stream processor applied cross-sensor correlation rules in real time — detecting composite conditions that involved readings from multiple sensor types — and routed alerts to a single alerting platform. Historical data was written to a time-series database with a query layer that allowed analytics tools to access all sensor histories through a consistent API regardless of original source protocol.
Results