United Parcel Service Boeing 767 34AF incident

This loss of required air traffic control separation resulted from the LN/LS controller issuing a landing clearance to UPS357 to a runway occupied by N900QS in position on the same runway. This resulted in UPS357 overflying N900QS by about 475 feet. Contributing to the error was the local controller's non-compliance with traffic advisory procedures, and insufficient requirements for use of memory aids to track arrival and departure operations.

When issuing a landing clearance to UPS357, the local controller did not ensure the runway was clear as required by FAA Order 7110.65, Air Traffic Control, paragraph 3-10-3, Same Runway Separation and, the local controller did not advise UPS357 that an aircraft was in position on the landing runway as required by FAA Order 7110.65, paragraph 3-10-5c, Landing Clearance.

Contributing to this incident was the lack of required memory aids for arrivals. Miami Air Traffic Control Tower procedures required controllers to use memory aids to indicate that an aircraft had been directed to line up and wait (LUAW), when an aircraft was issued a departure clearance, and to indicate that a runway was occupied by an airport vehicle. However, flight progress strips or memory aids were not required for arrivals and arrival flight progress strips were not used. The data available via the radar presentation on the tower display workstation (TDW) was the only information used for arrival sequencing. While the LUAW memory aide is beneficial to prevent a controller from issuing a takeoff clearance for a runway that is occupied by an aircraft in LUAW position, it did not prevent the controller from issuing a landing clearance to UPS357 nine seconds after he had directed N900QS to LUAW. There were no visual aids utilized to indicate an aircraft was on the runway to which an aircraft had also been cleared to land. According to controller interviews, the "rule of thumb" or unofficial procedure for arrivals to a runway occupied by an aircraft in position was that if the arrival reached two miles and the aircraft in position had not started departure roll, the arriving aircraft would be sent around. However, there were no visual clues or memory aids to remind the controller that an aircraft was in position on the runway except where a controller may decide to write down call-signs of arriving aircraft at the control position.

Verbatim NTSB analysis and probable cause from the NTSB dataset