Disclosed is an optical disk handling system (24), called an autochanger, having two control systems (1504, 1502) to provide the six motions necessary to move optical disk cartridges (43) from storage holding cells (40) to an optical drive (26). Whenever the mechanisms are being moved, the control systems monitor the progress of the movement, and should the control systems detect an error condition, a state update flag is changed to prevent further updates of the state of the system (176). With state updates disabled, the control systems (1504, 1502) will stop all mechanisms and prevent further movement of the mechanisms. The software that calls the control systems (404, 406, 408, 410), however, is unaware of this change in the state update flag, and continues to process until the operation appears complete. When the operation appears complete, an error recovery software function (412) is called. When the error recovery software (412) detects that the state update flag is in an error condition, error recovery is initiated. The type of error recovery performed is based on the state of the machine (176) at the time the error condition was first detected, since this state has been saved by changing the state update flag. The first part of error recovery (422) attempts to put the mechanisms into a known state, based on the mechanisms positions at the time of the error. After the mechanism's state has been established, error recovery movement is started, based on the operation that was being performed at the time of the error, and the state of the machine (176) at the time of the error. The type operation being performed is saved in non-volatile memory, so that the type of operation will be saved should electrical power fail during the operation. Therefore, after power is restored, power fail error recovery can perform error recovery based on the type of operation being attempted at the time power was lost.