What come interesting next is the ability of the machines to request services from specialist modules hosted at the edge or cloud to establish its own level of KPI. For example, a call to the IBM Watson for calculation of its component degradation level before triggering necessary service request for maintenance, a trigger to the AGV route optimization engine to reduce the transfer cost or time, or even a trigger to the OEE module for self-benchmarking. This type of service request will increase the “bidding” chance of the machines thereby optimizing the use of the machine in real-time. On top of that, material movement will be planned by the machines thereby establishing a localized strategy to manage the inventory level based on risk-cost balance. This risk can also be jointly managed within a cluster to establish efficiency in inventory level.