Towards the implementation of AGV's (Automated Guided Vehicles) complex navigation tasks within the dynamic and hazardous environment of upcoming smart factories the developers challenges reliability and maintenance problem. While the number of AGVs in shared space gets higher the risk of failure follows the trend and both the users and developers need an adequate tool to monitor and diagnose all applied AGVs in real time. We have prepared a decision network to support the AGVs missions by real-time internal diagnostic to prepare more reliable AGV solutions in smart factories. The internal diagnostics monitors the sensors/inputs signals and internal communication channels to take decision on further actions to perform within the given environment . The main task for the decision network is to support the control system or human operator with possible decisions to take to avoid system failure.