Project Background
At Panevo, I built from a scratch the first version of a novel real-time data platform for lean manufacturing, used in fast-moving production lines. Sensors on production lines send data every second to our cloud-based systems, and live metrics are shown in real-time to operators. When the line stops, a notification is shown to the operator so that they enter the cause of the downtime, allowing the company to collect very detailed information about the efficiency of their manufacturing environment.
Under the hood, data is processed through different backend services (MQTT broker, Kafka, Faust...) with real-time analysis, before being sent to storage (PostgreSQL) and for display in dashboards (Websockets). The platform runs on a cluster of virtual machines in Azure, orchestrated with Kubernetes and Docker Swarm (we are transitioning services from Swarm to AKS and this work is still ongoing).
Key Features
- Real-time dashboards
- Notification system with automatic alerting and escalation
- Complex reports for data exploration
- Many background jobs and services for data preparation and automatic reports
- Flexible REST API
- Role and Permission management system
- Enterprise Compliance and Security (MFA, SOCII...)
Technologies Used
