DFXBOT / Application guide
Can Industrial IoT Work Without Internet? Local, Cloud and Offline Workflows
Separate an internet outage from a local network outage. Learn what industrial IoT can do locally, where cloud access is needed and what to test before a pilot.
Internet access and a local network are different dependencies
A factory device can communicate over a local network without using the public internet. That does not mean every connected application works offline. The complete path matters: sensor, gateway, local network, application, database and any business-system interface. Identify where each function actually runs before making an availability claim.
What can continue locally
A local application may continue an agreed measurement or operator-guidance workflow when the local server, power, network and required data remain available. A cloud-dependent screen or decision service may stop when its internet connection is lost. A barcode job that has already been supplied locally is a different case from a new task that must be fetched from a remote system. Confirm each required behaviour in the proposed design.
Stored data is not automatically synchronized data
Store-and-forward operation needs explicit buffering, timestamps, capacity, retry and reconciliation rules. Specify what is retained, for how long, and what happens when storage is full. On reconnection, the application must handle duplicate messages, late events and conflicts with records that changed elsewhere. Do not assume a disconnected device can invent a valid business transaction.
Plan a controlled fallback
Define which actions are allowed during an interruption, which must pause, and how operators can recognize stale information. Keep a practical procedure for checking material and recording exceptions. The appropriate fallback depends on the consequences of using an outdated quantity or instruction, not simply on whether a device still has power.
Test the failure modes separately
Test loss of public internet, loss of the local network, an unavailable application server and loss of power as separate cases. Observe measurements, operator instructions, data retention and recovery. Use a controlled test environment and agreed procedures; avoid disrupting a live production process to discover the limitations.
Choose deployment around the workflow
Cloud and on-premise deployment each require clear responsibilities for access, backups, updates and support. For a stock-monitoring or guided-picking pilot, document the functions that must remain available and the interfaces required. Offline behaviour must be specified and validated for that installation; this article is not a promise that every SLIM or Pick To Light configuration supports all of these modes.