We develop an IoT edge device for collecting and processing industrial IoT data. The device supports dozens of connectivity protocols (e.g., MQTT, Kafka, REST) and backend systems (e.g., Hadoop, Cassandra, SQL databases). Fog computing applications - from data compression to machine learning - can be deployed on the device as docker containers. Computing thus takes place "on the edge" where the data is created, enabling fast reaction times and avoiding the security implications of sending data to cloud systems.
The devices can be centrally managed through a modern web application that provides monitoring (device status, memory, cpu, security notifications), planning and roll-out of OS and security updates and management of the deployed fog applications (container management, roll-out of application updates, ...). Intellectual property contained in customer-specific fog applications is protected by strong security mechanisms where on supported hardware, the TPM chip can be used as a trust anchor.