| dc.description.abstract | Introduction: The purpose of this document is to construct a recommended course of action in the next year for Garuda Operations Control in its efforts to upgrade its information systems technology. The process of installing new technologies is not one that can be done quickly or easily. It is also not one that can be accomplished by simply purchasing new software, even if that software were to exist. Rather, the process of upgrading technologies must follow a carefully planned and designed path. Among information systems specialists, the process is often referred to as the Systems Development Life Cycle (SDLC). The scope of an SDLC can vary. For airline operations control projects, the scope of the SDLC process is large. It involves many people, both internal and external to the organization. It requires the establishment of a Systems Development Team with membership from several units of the airline to direct the project and to resolve problems. It (ultimately) involves a substantial resource commitment, typically on the order of $2,000,000 to $3,000,000 in development funding. It involves a number of tasks that need to be performed as part of the development effort. And the project typically takes a number of years to implement. Failing to follow a proper Systems Development process may lead to a number of risks, such as: e The new system may not meet the user's needs. e The acquisition of unnecessary or inappropriate hardware. e The acquisition of insufficient software, or software that does not allow the airline to grow or handle future expansion. e Software that may be inadequately tested and may not meet requirements or expectations. One way to look at systems development is to divide it into six phases: Phase 1 - Analyze the current system Phase 2 - Define new system requirements Phase 3 - Design the new system Phase 4 - Develop the new system Phase 5 - Implement the new system Phase 6 - Test and evaluate the system's performance and its ability to meet the user's requirements During the last year, MIT/FTL staff have been working on Phase 1. The results of our analysis of GA's current system have been documented in a separate report by Michael Clarke and Yudi Naryadi entitled "The Airline Operation Control Centre: An Overview of Garuda's Operation Control (EM) at Cengkereng", which was recently submitted to GA. Perhaps more work needs to be done in Phase 1 by GA internal staff after GA has reviewed our report. For example, it might be wise to: a) Evaluate the sources of all data needed to support operations control. b) Document the flows of these data as EM goes about solving various operations problems, or resolving irregular operations. c) Document the information needs which are not currently available. d) Review current EM policies and procedures to obtain suggestions for improvement. However, it is the next two phases in the SDLC process (Phase 2 - defining the new system requirements, and Phase 3 - designing the new system) for which we now need to turn our attention. Within the next year of the cooperation between MIT and GA, there are a number of tasks that can be accomplished to complete these next two phases. What follows is our suggestion for what should be accomplished within the next year. 2. Suggested steps for the next year of cooperation between MIT and GA Operations Control Step 1 - Establish a Systems Development Team. The very first step that should be taken is the establishment of a team of individuals from both within GA and external to GA. The mission of this team would be to oversee the development effort: direct all activities; approve all decisions; make recommendations on the design of the new system; and resolve problems that occur along the way. The team should consist of personnel from: e Operations (EP, EM) e Flight Dispatch, Navigation (EA, ON) e Operations Control Center (OCC) e Maintenance (MCC, MP) * Crew Planning (OB) e Airport Operations (KO) e Information Systems (DX) The team should have a leader from within GA, and MIT/FTL staff would act as "consultants" to this team. Step 2 - Complete Phase 2 of the System Development Life Cycle. In the second phase of the SDLC, we need to scope out the requirements for the new system in enough detail so that both the computer systems developers and the users know exactly what the new system is going to do and how the system is going to do it. Needless to say, these requirements should solve the problems identified in Phase 1. The requirements should identify the user's needs (what the system will do) as well as the hardware, software, and data needs. This phase concludes with a system requirements report. Step 3 - Configure and install the computer hardware and networking technology that is necessary to allow personnel to electronically communicate and interact with one another, make good use of existing Operations Control systems, and to establish reliable access to all necessary information/data. The design of the hardware and network configuration is not a trivial task. Questions need to be answered: e What would be the underlying operating system: UNIX, Windows NT? e What hardware will the system run on: 80486 PC's or UNIX Workstations? e What client - server architecture is optimum? e What local area network is best: Ethernet, Token-Ring? * What media: Twisted-Pair, Co-ax? e How is the network to be connected to the mainframe and other systems? e What communications and network software is needed? It is planned that the installation of this hardware and software will be incremental and evolutionary. GA can initially procure just a few workstations and connect them up on a local area network. This "test cell" of computers will allow GA to gain some experience with the new hardware before making a more substantial commitment of resources. In addition, this step will allow EM personnel to become familiar with the new computer hardware before the application software is designed and installed. It will also allow EM personnel to communicate with each other through a local area network. In addition, the hardware and operating system software that is chosen should allow EM to continue to access and use current systems, even if those systems are on the mainframe computer or other workstations. At the same time, it should allow an evolutionary transition to better systems and software. Step 4 - Begin installation of a centralized Database Management System to hold the data items that are needed for effective Operations Control. Refer to the earlier proposal entitled "System Operations Control Database Development" written by Dennis Mathaisel in July 1995 for a more detailed discussion of this step. Configuring and installing an effective DBMS is not trivial. It is intended that an improved DBMS will be available on-line at EP/EM by transferring and updating data currently in other systems. Step 5 - Complete Phase 3 of the System Development Life Cycle. This third phase focuses on the design of the new system software before the software is procured or developed. The phase involves two main objectives: e To optimally design the new system. e To establish a sound framework of controls within which the new system should operate (basically, meeting the requirements). The completion of the design phase is marked by a couple of events: the team completes, organizes, and assembles the system design documentation; and a series of meetings/presentations are organized to present and review the design proposal. From an overall perspective, next year would be devoted to a year of assessment and design, combined with the installation of necessary hardware, operating systems, and local area networks. It would require a commitment from Garuda to purchase necessary hardware and LAN technology, as well as taking the first steps necessary to install a centralized DBMS. 3. Beyond next year... Once the above steps were completed, then GA can begin to acquire more advanced software to assist in planning and execution of Operations activities. The greatest mistake would be to acquire existing software packages before a thorough study and design was completed. A complete plan for developing a new operational system must be established first. Beyond next year, the basic steps would be as follows: a) Complete the construction of the centralized DBMS. b) Replace the ROC system currently in use in Operations Control with advanced computer-graphics displays on high-powered workstations that are connected on a local area network and connected with the mainframe computer. This step involves a transition to UNIX-based software. c) Then, and only after the above steps were taken, consider the introduction of automated decision-support models to solve specific problems that are encountered in irregular operations, etc. | en_US |
