Abstract: A system and method for predictive maintenance of multifunction peripherals includes a data collection agent configured to receive and store simple network management protocol (SNMP) data including metered data from one or more multifunction peripherals, and a service prediction system configured to selectively perform linear regression analysis on one or more of the metered data to predict one or more service events for at least one multifunction peripheral based on the linear regression analysis. Notifications for each predicted service event are sent to service technicians associated with the associated multifunction peripheral. Service technicians can preemptively perform service and maintenance functions on the multifunctional peripherals before error or faults occur on the multifunction peripherals.

Abstract: A system and method for predictive maintenance of multifunction peripherals includes a data collection agent configured to receive and store simple network management protocol (SNMP) data including metered data from one or more multifunction peripherals, and a service prediction system configured to selectively perform linear regression analysis on one or more of the metered data to predict one or more service events for at least one multifunction peripheral based on the linear regression analysis. Notifications for each predicted service event are sent to service technicians associated with the associated multifunction peripheral. Service technicians can preemptively perform service and maintenance functions on the multifunctional peripherals before error or faults occur on the multifunction peripherals.

Abstract: A system and method for predictive object identifier (OID) field identification for multifunction peripherals includes a processor configured to walk the managed information base (MIB) of a multifunction peripheral having unidentified OIDs, and save simple network management protocol (SNMP) data associated with the unidentified OIDs to a database. A machine learning process compares data associated with the unidentified OIDs with training data derived from prior training of the machine learning process to predict one or more candidate OID field types for one or several unidentified OIDs. One of the candidate OID field types can be selected by a user as the predicted OID field type, and the selection can be used to improve the training of the machine learning process. The machine learning process can be trained using data derived from walking the MIB of a plurality of multifunction peripherals having known MIBs and OIDs.