Sir

In the News Feature 'Phoenix: a race against time' (Nature 456, 690–695; 2008) you report on a problem that stopped the doors to the ovens on the Phoenix spacecraft's Thermal and Evolved Gas Analyzer (TEGA) instrument from opening fully. You note that the University of Arizona team responsible for TEGA noticed the door interference problem during engineering tests and sent revised designs to Honeybee Robotics of New York, but that Honeybee Robotics sent back new parts using the “original flawed designs”.

We at the University of Arizona are concerned that Honeybee's responsibility for the TEGA door opening has been overstated. As the dust-cover manufacturer, the company played a very minor role in the overall TEGA construction. Most importantly, it was not responsible for establishing that the TEGA doors functioned properly after the dust-cover assembly was installed. Unfortunately, because the door-opening mechanism is a single-use actuator, which is buried deep within the instrument, we on the instrument team could not test it on the flight unit without a time-consuming process of dismantling and rebuilding the system.

In light of the publicity given to the door problem, we would like to set a few of the facts straight. When the door-opening problem was first observed during testing of an engineering prototype, we documented the needed change, along with some other changes, in an engineering drawing that was a modification to an earlier Honeybee drawing. Our drawing had more than a dozen changes on it, all but one of which were well documented with new dimensions called out. However, the change that related to the issue with the doors, although drawn properly, was not explicitly called out as a change in this way.

We should have asked to review the final drawing before metal was cut, but we did not. As we could not test the operation of the doors in the flight unit, we should have checked the parts when they arrived, but we did not. I recognize that my group at the University of Arizona operates a bit more informally than is current practice in the aerospace industry; this has served us well in the past in keeping costs down, but entails risk. We should have caught the problem and we didn't. In the end, the buck stops with us.

There is nothing associated with this event that changes our full confidence in Honeybee's competence. We have had an excellent working relationship with the company on this and several earlier projects, and Honeybee would clearly be our vendor of choice for similar projects in the future.