This is a summary of a poster session presented at the 1993 CAUSE Annual Conference held in San Diego, California, December 7-10. It is the intellectual property of the author(s) and may not be published without permission. Disseminating the material otherwise is permitted, provided that the author(s) and the original presentation of the 
paper at CAUSE93 are acknowledged. 


                           DOWNSIZING?
                         GO ALL THE WAY!


Joel Hartman, 
Associate Provost for Information Technologies and Resources
Bradley University
joel@bradley.edu
(309)677-3100

Steve Patrick, 
Director, Computing Services
Bradley University
steve@bradley.edu
(309)677-2950


BRADLEY UNIVERSITY COMPUTING INITIATIVE


PURPOSE?
     
Examine future administrative computing needs and develop an 
architecture to meet those needs.
     
Key elements:

  *  Universal Access to Computing

  *  Network Access to Information
          
  *  Network Access to a variety of Computing Solutions
          
  *  Migration from mainframe to networked computing
          
  *  Quality, Reliability, and Service
          
  *  Executive Information Support
     
  *  Show how to build administrative systems that are 
     powerful yet easy to use.

Three main constituent groups:

  *  Faculty and Students - Teaching, learning, and scholarly
     activity
     
  *  Administration - Operational Systems, clerical,
     professional, and middle management levels.
     
  *  Executives - Decision making support and planning,
     Presidential, Vice Presidential, Dean, and Department
     Chair levels.
                                
An Individual may be a member of more than one group
This presentation focuses on the needs of administrative 
computer users.


ADMINISTRATIVE SYSTEMS

Values important to administrative users are:

  *  Reliability
     
  *  Consistency
     
  *  Service
     
  *  Cost Containment
     
  *  Ease of use

History
     
Bradley University, like the vast majority of 
educational institutions, used a single mainframe computer to 
support our administrative applications.  The mainframe 
computing approach evolved at a time when computers were 
complex, costly, and required expert support.  We can now 
support our administrative applications on smaller, less 
expensive, independent computers (distributed computing).
     
     
COMPARISON OF DISTRIBUTED AND MAINFRAME COMPUTING      

Financing

Mainframe Computing requires buying a "3/4 million dollar" 
computer every 3-4 years. Distributed computing requires 
buying or upgrading a smaller computer every year (our goal 
is level annual capital expenditures).  We can extend the 
useful life of our distributed computers by using a parallel 
processing architecture. Small computers are experiencing a 
50%-100% increase in price performance each year.  This 
suggests that a lower "constant dollar" expenditure will be 
more cost effective than the intermittent high cost 
acquisition of a mainframe computer.  We can deliver more 
capability for the same or lower cost.
     

Responsiveness
          
Replacing a mainframe affects all offices at the same time.  
This prevents us from easily responding to changing needs. 
Changing a distributed processor affects a single 
administrative function.  We can concentrate our resources on 
that change.


Performance

The addition of new applications or more users affects all 
other users.  Mainframes typically have acceptable 
Performance upon installation followed by a slow but steady 
degradation until the mainframe needs to be replaced.
                                
                                
REENGINEER ADMINISTRATIVE PROCEDURES AND APPLICATIONS 

Reengineering is:
          
Using the power of modern information technology to
radically redesign business processes in order to
achieve dramatic improvement in performance.

Re-examine all of our basic assumptions about the way
we do things and rejecting those that do not fit
today's technological capabilities.

Redesigning work processes based upon new assumptions.

"Thinking out of the box."  Refusing to be limited by
traditions of the past.
     
Leverage technology to achieve an order of magnitude
increase in productivity.
     
Examine all aspects of administrative functions being
automated to look for opportunities to provide better
service and reduce costs. 

Optimize procedures rather than merely automate existing 
manual systems.

Eliminate unnecessary paper.

Distributed Computing

Performance can be tuned to an application.  We can fine
tune our purchasing decisions to meet the current as well as
planned needs of a particular office.

Addition of load in one area affects performance in that
unit only.  Other computer users are not affected by the
changes of a different application family's needs.

Performance is controllable by each individual office.
The users can determine priorities and are in greater
control.

Performance increments cost less on smaller computers
making it easier find the funds to upgrade distributed
computers to meet growing needs.

Open Systems

Computing environments supported by a variety of computer
vendors.

Process any application on any computer

Survivable Systems (Multiple computer room philosophy)

Maintenance friendly (In-house maintenance)

No more massive administrative conversions due to a
mainframe change.


ADMINISTRATIVE FAMILIES OF APPLICATIONS

Student Family
     
     Admissions
     
     Registration
     
     Records
     
     Financial Aid

Financial Family
     
     Accounting
     
     Purchasing
     
     Payroll and Human Resources

Development Family
     
     Alumni Records
     
     Fund Raising
     
     Telemarketing
     
     Campaign Management

Administrative Workstation Configuration:

     MS-DOS 386 or better.
     
     Connected to high speed network (10 Million BPS).
     
     The administrative environment is much more homogeneous 
     than the academic environment.
     
     Higher level workstations will be used by executives,
     managers and professionals.  The new client-server
     architecture provide seamless integration between
     microcomputer applications and administrative data.  For
     example, administrators will be able to access
     administrative data from within a spreadsheet.
     
     

SOFTWARE SUPPORT
     
Make vs. Buy Economics
          
We need to look at the long term impact of our  software 
decisions.  Administrative software systems provide the 
foundation for our information systems and should have a life 
span of more than 20 years.  This will encompass several 
hardware changes.
          
  *  Make Advantages
          
  *  Lower initial "out of pocket" cost
          
  *  Customized to serve our environment
          
  *  We must change our procedures
          
  *  Buy Advantages
          
  *  Faster implementation
          
  *  More responsive maintenance
          
  *  Lower total cost
          
  *  Easier to determine in advance the system's
     capabilities.
          
  *  Meets industry standards (FASB, GAAP, etc.)

Administrative Software Design Strategies

  *  Purchase Core Operational Systems
          
  *  Payroll and Accounting
     
  *  Develop Strategic Applications
          
  *  Student Records
          
  *  Development and Alumni
          
  *  Admissions
     
  *  Insure that Data are Stored in Relational Databases
          
  *  Ease of modification
          
  *  Ease of ad hoc inquiry
          
  *  Integrate into ad hoc applications - We will surround
     the production database with user-controlled
     applications to provide true added value to our
     production systems.


ISSUES
     
Administration
          
Administration (systems, special forms and high volume
printing, backups) will be performed by Computing Services.
Data entry and scheduling of administrative applications will 
be performed by the user office.
     
Security

The network will be split into an Administrative and an 
Academic network.  Barriers will prevent access to 
administrative computers from residence halls and student 
labs.
          
Modern databases can provide item-level security to
individuals for both update and inquiry.

Reliability

Reliability can be enhanced by distributed computing.  Using 
"open systems" located in different buildings enhances 
reliability because we can move applications to maintain 
service in the event of a disaster.  Loss of a distributed 
computer affects only users of that computer rather than all 
users.
          
          

ADMINISTRATIVE COMPUTING IMPLEMENTATION STAGES 

Phase 1 - 1992-93
     
Advancement - purchase hardware and develop the system in
client/server environment.  Began development in June 1992.
System implemented in June 1993.

Phase 2 - 1993-94
     
Financial System - purchase software and hardware.  Software
and hardware purchased in June 1993.  Payroll implementation
date January 1994.  Remaining financial systems June 1994.

Phase 3 - 1993-1995
     
Student System - Began development in client/server 
environment in June 1993.  Planned implementation summer
1995.
     
Admissions System - Planning to begin development in January
1994, implementation in summer 1995.

Phase 4 - 1995
     
Remove mainframe - before we remove the mainframe we must
migrate academic computer users to a different computing
platform.

The transition to the new environment will be a prolonged, 
but controlled and orderly process.  We will concentrate our
development resources on a single group of administrative
functions during each phase.  This will provide a relatively 
fast transition for each office to the new environment.  
During the transition, users will move from the mainframe to 
distributed processors.  The removal of computing load from 
the mainframe will provide better service for the remaining 
mainframe users.
                                
                 Hypothetical Capital Budget
           For Continuing the Current Environment
                                                          
                       FY 93     FY 94     FY 95     Total
                                                          
Computer Hardware   $240,000  $240,000  $240,000  $720,000
                                                          
Hardware Maint.      $20,000   $23,000   $26,000   $69,000
                                                          
Software Maint.      $20,000   $22,000   $24,000   $66,000
                                                          
Total               $280,000  $285,000  $290,000  $855,000


To continue to support all administrative computing on a
mainframe, we will need to expand to a larger system.  Our
current system does not have the capacity to support modern
database applications on top of its current and projected 
load.

The resulting system will be faster and provide better 
responsetime, but will be the same type of system we are now 
using.  Every three years, growth in demand will mandate a 
mainframe computer upgrade.  At the end of the planning 
horizon we should be ready to procure a new system.
                                
           Hypothetical Capital and Operating Budget
                For the Proposed Environment
                                
                                                             
Project                Advance-   Finance   Student     Total
                           ment              & Adm.
                                                             
Central Hardware        $21,000   $32,000   $76,000  $129,000
                                                             
User Hardware           $59,000   $27,500   $32,000  $118,500
                                                             
Applications Software             $26,000             $26,000
                                                             
Training & Consulting   $20,000   $10,500   $40,500   $71,000
                                                             
Total Cost              $90,000   $96,000  $148,500  $334,500
                                

CONCLUSION

The Computing Initiative is a tactical plan to achieve the 
goals of the strategic plan.  The major benefits of this 
initiative are:
     
  *  A plan to take Bradley University to the year 2000 and
     beyond.
     
  *  Move from a "closed" single-vendor mainframe environment
     to a distributed "open" vendor-independent environment.
     
  *  Migrate administrative systems from an obsolete, flat 
     file organization to a modern database environment.
     
  *  Improve the systems development technology from a COBOL
     based environment (1960's software technology) to fourth
     generation computer languages (1990's software 
     Technology).
     
  *  Provide systems that can rapidly respond to changing
     needs.
     
  *  Provide administrative systems that are easy to modify.
     
  *  Increase the survivability of our systems.
     
  *  No mass administrative conversion.


Date Revised: 02/24/94 (bdz)