The Titan IV B Configuration Vehicle - Case Study Example

SUMMARY OF MISTAKES AND LESSONS LEARNT FROM THE TITAN IV ACCIDENT Background The Titan IV B configuration vehicle was launched into space with Titan centaur upper stage on April 30, 1999. The mission of the Titan IV was to place Milstar Satellite in a geosynchronous orbit. This was not accomplished. The satellite was placed in low elliptical orbit due to mechanical problems encountered by Titan IV. The Accident Investigation Board and Safety Investigation Board were appointed by the Acting Secretary of US Air Force to concurrently carry out a dual track investigative process on the Titan IV B-32 mission failure. An Engineering Analysis Team comprising personnel from Lockheed Martins Aeronautics, Air Force and Aerospace Corporation conducted a technical analysis which informed the opinion statement of the two supervising boards. The investigation concludes that the inability of the failed process to detect and correct human error in the manual entering of roll rate filter constant. More technically and specifically, the following major mistakes were identified: 1. Software Development a. The software was not well defined, documented and/or well understood by no single player among the multiple players involved in the mission process. b. No adequate formal checks in the process to allow the creation and updating of database for operating the Titan IV. c. The Titan Iv software was designed with single point error for mission critical data. There was no room for the system to troubleshoot itself by detecting, correcting and/or automatically using alternative permutations of critical data. 2. Testing, validation and verification a. The independent verification and validation program approved was not functional. The program was unable to detect and correct the wrongly entered roll rate constant. b. There were no formal processes for validation of the I1 filter constants or monitoring attitude rates. c. There was no direct communication among the responsible parties thus preventing the correction of problems or clarification of not-too-clear procedures during testing and maturation of the Titan IV. 3. Quality/mission assurance a. There was no comprehensive surveillance plan to cushion factors (personnel reduction and transition from oversight to insight roles) external to the operation environment of the Titan IV. The responsibilities of the supervising government agencies as the take the insight role were not well defined. The process and the software program were not well understood thus the transition from oversight to insight was not well implemented. From the findings of the investigation boards, the following lessons were learned: 1. Success of this type of project depends largely on accuracy and precision (in this case, error in decimals was one the major causes of the Titan IV mishap). These attributes can only be achieved if the whole process is adequately formally documented. In the case of the Titan IV project, however, formal documentation of the process was abhorrently casual. There was too much dependence on individual expert’s effort and work, with no tracking and recording of procedures and actions taken. This was manifest in the lack of adequate formal documentation of software development and its operating manual, and procedures for creating and updating database. For such a precision-dependent process, there is no room for guess work or trial-by-error. 2. It was obvious that an expertise (knowledge/skills) vacuum was created due to poor contractual/planning. In the design of technology-based projects like Titan IV, the experts involved must be retained from conception to completion (successfully testing of new technology). It does not matter if these experts are initially hired by different firms contracted by a government agency. The supervising agency must ensure that they are either assimilated into the main service or the responsible government agency works out a special retainership that will keep them together, even for a post-mortem should in case the project turns out to be unsuccessful. This therefore implies that supervisory agencies must carefully design contracts that will accommodate the interests of employers (parent companies) of these experts while ensuring continuity. 3. Related to the above, lack of a master surveillance plan contributed to the lost of expert knowledge and break in line of operations from design to testing. There was no plan to cushion the effects of external factors (personnel reduction in the supervisory agency and transition in role). Although, there was a program office created for the Titan IV project, there was no permanent official of the supervisory government agency to work with the companies hired at different stages (from design to testing to launch) and components of the project (since different component parts were produced by different companies). It was clear that the supervisory directorate did not make adequate arrangement to backup expert knowledge that will successfully guarantee smooth transition from design to testing to launch. Few of the experts that designed the plan and understood the process remained after the ‘completion’ of the project. Thus while there was fragmentation of responsibilities of companies managing different experts (involved in the production of different components of the Titan IV), there was no provision for a platform for direct communication and interface among these companies. Confusions and uncertainties were therefore not cleared. It shows clearly that projects can not be successful if they are not sustainable. Sustainability, here, requires adequate management (by adequately formal documental) of technical knowledge of the design, test and final launch operations (including maintenance). It also requires constant direct interaction between companies contracted for project and interaction with the responsible government agency(s). 4. The validation and verification was done by one person, who had limited knowledge of the whole process. For projects that depend on high precision and accuracy there must be multi-layered formalized validation and verification systems. This would give room for multiple-checking and resolution of mistakes. Conclusion Mistakes identified at individual and organizational levels are largely rooted in the casual documentation of procedures from design to testing stages. Also, the software was not well designed to allow system troubleshooting in case of failures due mistakes in entering system critical data. The whole process was not well understood by those involved. Beside there was no direct communication and interface among the experts (working for different companies and government agencies) involved in the process. Gaps and vacuums were inevitably created, and with no well-documented procedures trial-by-error became a lead out of the dark. The Titan mishap can also be attributed to emphasis on task completion rather than task implementation. Verification and validation were not meticulous thus quality assurance/surveillance was compromised. There was no plan for smooth transition of supervising government agencies from oversight to insight role. Read More