The Hubble Telescope Quality Management Failure: Who is to blame?

"The difficulties with the Hubble and the space shuttle program are very disturbing and, without judging the cause of these two incidents, they have served to reaffirm my strong belief in the need for established quality assurance procedures and strong effective program management by NASA."

These words, vocalized by Al Gore 3 months into the project, echoed what NASA already knew. The Hubble telescope had a serious flaw that had caused the mission to be deemed a failure. But how could it have? Almost 15 years and $2.5 billion had been pumped into the mission. There’s no way a project of this magnitude could have any blemishes could it?

The problems with the Hubble went beyond the layer of a simple technology issue. It exposed a much deeper issue, that of a serious lack of quality management and leadership. The quality of images returned by the telescope were not as expected and the images themselves were blurry. To add to this, the primary objective of capturing images of far off objects was not met. Investigation of the problem revealed that the primary mirror, created by Perkin-Elmer, was ground to the wrong shape. The Hubble Space Telescope was launched with a spherically aberrated mirror.

                                                         Images before and after the correction of the mirror

Apart from the spherical aberration of the mirror, the mission also had other problems like , flopping solar-energy panels, velocity-measurement system failure, chemistry-of-celestial-objects systems failure and three faulty gyroscopes. The cost of the project had cost overruns of about 463% and a third of the memory banks were faulty. During the prototype phase, Perkin Elmer had assigned its best engineers to the job of building the prototype for the Telescope. However, the same was not done so while the final mirror was being developed. As a result, the shape of the mirror was off by  2.2 micrometers (Yes! that small!) and was thus spherically aberrated. A null corrector is a device used to test large aspheric mirrors and achieve a precise shape. Initially, the company was using conventional null correctors while a custom built null corrector had been used for the final production of the mirror to try and achieve better quality. Paradoxically, this null corrector had not been assembled correctly itself, which led to the flaw. The flawed null corrector indicated that the mirror was perfectly shaped. However, final tests performed with conventional null correctors did expose the aberration in the mirror. The management, however, chose to overlook these results and decided that the tests of the custom built null corrector, which had indicated no errors, were more precise than the conventional mirrors.

An interesting observation here is that NASA and Perkin-Elmer were going through a tough period as partners. Due to the cost overruns and scheduling conflicts, relations between the two organizations had strained. As a result, Perkin-Elmer did not inform NASA about problems immediately, but at their convenience and also did not concentrate all their best minds on the final mirror of the Hubble. Lack of communication had also led Carl Fuller, the quality assurance manager at NASA, to ignore the quality of one piece – the mirror, since the two organizations had not coordinated with each other during the final tests. Apart from the billions had to be spent to correct this flaw, both NASA and Perkin-Elmer faced public humiliation for lack of quality management and harmonization.

Who is to blame for the Hubble Telescope disaster? Perkin Elmer for trusting an object they perceived to be more precise than its predecessors, or NASA, for choosing Perkin Elmer in the first place?

It is essential for organizations to pay close attention to details, be it a project of any magnitude. Delivering products that conform to the quality requirements is a no brainer. The Hubble mission goes to show how a failure of TQM can cause major damages to corporations and how even ultra large scale projects are not insulated from quality management breakdowns.

References:

http://www.cio.com.au/article/420036/what_went_wrong_hubble_space_telescope_what_managers_can_learn_from_it_/

http://www.qualitydigest.com/oct06/columnists/jharrington.shtml

http://www.ssl.berkeley.edu/~mlampton/AllenReportHST.pdf