Knowledge based resource management

#This invention relates to maintenance systems, and in particular, to the use of a field replaceable unit fencing action within a multiply redundant, fault tolerant customer subsystem to dynamically manipulate individual functions within the subsystem and maximize resources which are available to the customer, while the maintenance system reconfigures around failures and processes concurrent diagnostic submissions, IML (initial microprogram load) protection, and `hot-pluggable FRU` insertion.

#Electronic device manufacturing techniques make it possible to incorporate a significant amount of functionality into a very small physical area. Customer systems now contain a great many functions on very few field replaceable units. Even though the reliability of individual electronic devices has improved, increased device density and functionality often increase the risk of catastrophic failure from any fault source. It is therefore often desirable to use redundant critical functional elements in the customer system.

#However, the use of a multiplicity of redundant functional elements creates an entirely new class of problems. In prior art maintenance systems, it is possible to remove or "fence" entire redundant functional elements from service by electrically disabling paths to or from such elements so that alternate equipment is available in case a primary equipment is lost due to a failure. The fencing capability alone does not permit efficient resource management. The devices contained within a fence must be precisely selected to avoid disabling elements that are required to maintain acceptable levels of system operation. In addition, only a single fence can be maintained, rendering diagnosis of multiple concurrent failures a difficult task. Furthermore, the fence procedure is applied retrospectively, attempting to recreate the unknown set of conditions that were extant at the time of the failure. It is often necessary to remove the entire customer system from operation in order to diagnose a failure within a single element. Thus, fencing is a valuable maintenance tool, but one with significant limitations in capability.

#FIG. 1 illustrates in block diagram form a distributed hierarchical machine initiated maintenance system;

#FIGS. 2 and 3 illustrate the knowledge based resource management system in further detail;

#FIG. 4 illustrates additional details of the isolate field replaceable unit function;

#FIG. 1 illustrates in block diagram form the hierarchical distributed knowledge based machine initiated maintenance system. A plurality of customer equipment 40-42 are illustrated connected to a central maintenance system 1 via corresponding communication links 50-53. Included in each customer equipment 40 or adjunct thereto is a machine initiated maintenance system 10 which functions to communicate with central maintenance system 1 in order to provide instantaneous and efficient error identification reporting and collection. Each customer equipment 40 can be directly connected to the central maintenance system 1 or, alternatively, a plurality of customer equipment 41, 42 can be connected to a regional maintenance system 30 and a plurality of the regional maintenance systems 30 can be connected to the central maintenance system 1. The maintenance system is hierarchically arranged with local customer equipment 40-42 being serviced by its associated internal maintenance system 10 and a pool of customer equipments 41, 42 and their associated maintenance systems 10 being serviced by a corresponding regional maintenance system 30. A pool of regional maintenance systems 30 are then connected to the central maintenance system 1. The sophistication and complexity of each level of the maintenance system can be selected to correspond to economic factors, such as the installed worth of the equipment that is so maintained. Therefore, the local maintenance system 10 may be of lesser complexity than the regional maintenance system 30 since it must be replicated in each customer equipment 40. The regional maintenance system 30 can be of increased complexity since it serves a plurality of customer equipments 41-42 and views the data collected from all of these systems to thereby obtain a broader perspective on error conditions. Similarly, the central maintenance system 1 can be a sophisticated and expensive system since it serves all installed customer equipment 40-42 and supports all the regional maintenance systems 30. The central maintenance system 1 maintains records of all failures that have occurred in all customer equipment 40-42 and can therefore detect failure patterns using this data that are beyond the scope of the ability of the regional 30 or local maintenance systems 10 due to the limited pool of data available to these systems.

#Customer Equipment Architecture

#Customer equipment 40 is typically constructed of a plurality of field replaceable units (FRU) 70-75, each of which performs a designated, well defined function. Additional elements may be provided in customer equipment 40 that are not field replaceable units, in that a craftsperson can not simply repair or replace these elements. However, for the purpose of simplicity of description, all functional subelements contained within customer equipment 40 are designated as field replaceable units whether indeed they can be physically replaced by a craftsperson in the traditional field replaceable unit sense. The physical configuration of the customer equipment 40 is not particularly pertinent to the concept of this invention and the term field replaceable unit should not be construed as any limitation on the operational abilities of the subject system.