Martin Deploys Stackable on Ionos for Data Sovereignty
Waste-to-energy firm avoids US hyperscalers to secure critical infrastructure telemetry.
1Strategic Infrastructure Shift
Munich-based Martin, a century-old waste-to-energy firm, transitioned its telemetry operations to the German cloud provider Ionos. By deploying the open-source Stackable platform, the company aims to maintain full control over its critical plant data. This move specifically addresses concerns regarding the reach of the US Cloud Act on industrial infrastructure.
2Operational Data Scale
The company manages over 1,100 incineration lines, each generating 10GB of telemetry daily from hundreds of sensors. These sensors monitor combustion, emissions, and energy recovery metrics. Previously, this massive data volume was handled through decentralised, manual processes, which hindered efficient analysis and limited the firm's ability to optimize plant performance effectively.
3Sovereignty Through Architecture
Martin selected Stackable because it packages components like Apache Spark, Kafka, and Trino as Kubernetes operators. This architecture allows the entire data layer to be moved between clouds or on-premises without rebuilding. By avoiding proprietary software and licence fees, the firm ensures it can exit any cloud provider whenever necessary.
4Technical Implementation Details
The current pipeline streams sensor readings from edge devices via MQTT into a NiFi broker. Data is processed through Spark jobs scheduled by Airflow into Iceberg tables, then visualized using Superset and Grafana. Infrastructure is managed as code using OpenTofu, with Ionos providing the underlying Kubernetes cluster for the entire deployment.
5Avoiding US Jurisdictional Reach
Max Schönsteiner, head of R&D, prioritized providers without US parent companies to bypass the US Cloud Act. While Ionos operates under European law, the firm acknowledges that sovereignty depends on the entire stack. The company aims to ensure that no US-based software or equipment compromises their control over critical facility data.
6Predictive Maintenance Ambitions
While currently focused on data quality and curation, Martin is developing AI-driven predictive maintenance. The goal is to analyze core components like boilers and furnaces to improve uptime. Because availability is the primary metric for waste plants, the firm is carefully integrating physical-thermodynamic models with real-time measurement data for accuracy.
7Deployment and Scaling Timeline
The roll-out began in January 2023 and achieved full production status within two years. Currently, 35 incineration lines are connected to the new platform. Martin plans to equip as many lines as possible across its global sites, eventually packaging these advanced analytics as a service for its diverse plant operator clients.
8Practical Sovereignty Lessons
Martin’s deployment serves as a working model for enterprises concerned about hyperscaler dependence. By building an exit option directly into the architecture rather than relying on supplier promises, the firm secures its operational future. This approach demonstrates that true sovereignty requires controlling the entire technology stack, from hardware to software layers.
Original article: computerweekly.com — Martin Deploys Stackable on Ionos for Data Sovereignty