![]() |
|||
![]()
|
![]() |
![]() |
![]() |
7.4. MANAGEMENT APPLICATIONS OF EXPERT SYSTEMSMany aspects of management can be automated, resulting in a higher level of performance than is possible from any human manager or management team. The management tasks with the largest potential gains include dynamic workflow modeling and simulation for process improvement; automated workload management for allocating, scheduling, and assigning cases; automated case management for work requiring multiple specialists; automated quality assurance; and analyzing and forecasting trends in measurement data. Dynamic workflow modeling and simulation: All business concepts, practices, and processes can be documented and available online as text, process maps, supporting work system elements, and process simulation models. In addition, all relevant measurement data, as well as performance standards and employee compensation criteria, will also be available online, and will be integrated with the process models. The Management Information System (MIS) will be integrated with the related process models. Thus, all employees will be able to access, design, and test new process models through simulation by modifying the input variables. Modeling and simulation software will assist in improving the work processes. Automated workload management: Work that is informational in nature can usually be performed independent of location simply by transferring caseloads electronically between sites. To accomplish this, an enterprise-wide object-oriented MBR is used to structure, organize, and sequence the work. Then, specialized algorithms can be selected by RBS and applied to level workloads dynamically enterprise-wide. When corporate workload leveling is combined with sophisticated workload allocation, scheduling, and assignment algorithms, large gains in productivity and staff utilization are possible. Automated case management: Cases can be structured using MBR to represent several important variables in allocating and scheduling work: the amount, type, location dependence, and timing of receipt of the work; the quantity, skills, prior performance, and availability of staff; and the required completion date and accuracy. CBR and/or RBS can be used to optimize work assignments as well as workforce development. When workloads are high and inventories are backing up, the scheduler module will assign new work to the highest performing specialists available at the lowest appropriate skill and experience levels. When workloads are low, the module will assign or suggest to managers that they assign developmental tasks to employees that stretch their abilities. The module can also be used to plan future workloads and resource needs. Complex work requests may require processing from a variety of specialists, sometimes in a certain sequence. This system serves as the case manager, creating a virtual single point of contact for the client. For example, hiring an employee might include tasks such as ordering furniture and computers, getting security clearances and ID badges, performing orientation and initiation, and establishing payroll and benefits. This system also allows customers to request a completion date that can be negotiated by a RBS module based on user priorities if the requested date cannot be met. The system can reschedule other work assignments and renegotiate related completion dates to meet a truly urgent request. The system also sends notifications to customers when it forecasts that agreed-to completion dates will not be met. Automated quality assurance (QA): All work performed electronically can be automatically sampled and compared for quality. CBR software allows for the comparison of stereotypical cases against newly worked cases. Sampled cases that are very similar to stereotypical cases are assumed to be correct. Cases that are dissimilar are referred to managers and/or appropriate QA employees. New cases are reviewed for accuracy and method of solution. There are three possibilities: the new case is either incorrectly worked, represents a new type of work, or has a different valid or better solution method than exists in the case base. In addition to online QA, the evaluation module can also generate simulated test cases in certain situations. In real time, the evaluation module generates test cases or selects them from a case base in response to recent errors in certain types of cases, or when analysis of performance measures indicates a problem. Test cases are sent to employees when QA or the manager detects errors. Test cases can also be sent to employees to evaluate recent training or the need for remedial training. Analysis and forecasting of trends in measurement data: Standardized Management Information System (MIS) reports will provide raw and percentage change comparisons to the immediate prior period as well as the same period for the prior year for all requested levels. Exception reporting with user-set default settings or highlighted variances of specified sizes can be selected. Users will be able to customize the reports to suit their own preferences by changing a few parameters. Sophisticated user interface design will allow users to more easily browse, explore, or produce desired ad hoc requests. A variety of techniques are available for analyzing the MIS results. RBS can detect significant variances in historical measurement data, and MBR can forecast future trends and problems based on causal or empirical relationships between measures. Text generation AI software can explain the variances in English prose. Process simulation MBR models can be used to predict the downstream consequences of the variances. Time-series trend analysis software also will be used to forecast future performance. Information access and presentation will have functionality similar to that described for external environmental scanning.
|
![]() |
|
Use of this site is subject certain Terms & Conditions. Copyright (c) 1996-1999 EarthWeb, Inc.. All rights reserved. Reproduction in whole or in part in any form or medium without express written permission of EarthWeb is prohibited. Please read our privacy policy for details. |