The purpose of this EMPI program was to design an Automatic Coil Winding Station (ACWS) for winding Fiber Optic Gyro (FOG) sensor coils through the use of TQM, QFD, etc., followed by use of Taguchi an other statistical techniques to optimize the coil winding process. Four phases were involved: Process Definition, Critical Factor Identification, Variability Reduction, and SPC Implementation. Winding FOG coils is both difficult and fragile in that it is a quadrupole wind - as apposed to the conventional thread wind - compounded by the requirement for low tension precision, high-fiber packing density, and always risk of damage to the delicate fiber itself. The critical factor identification in the quadrupole winding process was reduced to fiber crossover - a significant detrimental influence on gyro performance - which, in turn, was closely identified with fiber gap control. The station was completed and deployed to the field where production coils are currently being wound. The ACWS not only lowered the required labor skill but succeeded in reducing the winding cycle time to 1 hour (from 24 hours) and touch labor time to 0.3 hours (from 24 hours) while improving the yield and performance through improved process control.