That approach is itself noteworthy. Space hardware is often designed and built direct from drawing board to finished product, using components and systems with some flight history. GP-B doesn’t have that luxury-nor does it have any margin for error. "This is not a straightforward engineering task," says John Turneaure, a co-principal investigator on the project. After working out the design concept, the team developed small portions of the hardware and demonstrated that each would work. Next, they put these pieces together in small groups for subsystem testing and finally assembled the subsystems in order to test the workings of the complete instrument.
In practical terms that means the team has had to build, test, and refine two prototypes of the GP-B probe that will never even fly in space. A scaled-down version of the probe, designed to test the gyroscopes’ performance in low gravity, is scheduled to fly on the shuttle in 1995.