This paper presents a novel fuzzy auto-tuning methodology to continuously adapt PID control actions without interrupting the normal process operation. New auto-tuning rules are introduced to schedule a gradual and guided activation of each individual P, I and D control mode in three adjustable functional control zones. An auto-scaling procedure has been incorporated to generalize the autotuning scheme to efficiently respond to any set-point change outside a pre-defined operating span. In contrast to existing auto-tuning algorithms, the proposed scheme is not an on-demand auto-tuning methodology and hence does not require alertness of an experienced engineer to initiate and supervise its initial operation in a separate commissioning identification pre-test. This interesting feature provides a new perspective on PID auto-tuning approaches. Performance of the proposed auto-tuning scheme is practically evaluated in a real pilot plant within a networked control system (NCS) configuration, realized by industrial Ethernet and Foundation Fieldbus technologies. An extensive series of test scenarios has been conducted to explore efficiency of the proposed auto-tuning methodology to cope with fixed and varying operating set-points under uncertain and variable network transmission time delays and external disturbance.