Simulator Evaluation of a New Cockpit Descent Procedure

An experiment was conducted to evaluate flight crew performance of the Descent, a new cockpit procedure designed to support the Descent Advisor (DA), one of the components in a new air traffic control advisory system called the Center-TRACON Automation System (CTAS). The DA predicts when aircraft will reach a specific waypoint on the arrival route, and presents controllers with clearance advisories designed to improve the sequencing of arriving aircraft. The effectiveness of the DA depends on the aircraft's descent trajectory: where it begins descent, what speed it maintains, how fast and at what altitude it crosses the bottom-of-descent waypoint. The Precision Descent allows controllers to assign these descent parameters to the flight crew. A Field Evaluation of the DA was conducted in Denver in 1995. Three separate clearances using standard ATC phraseology were used to support the cockpit descent procedure during this evaluation. The number and length of these clearances caused problems for both controllers and flight crews, causing readback errors, repeat requests and procedure misunderstandings. These observations led to a focus group meeting in which controller and pilot participants in the 1995 FE assisted in the redesign of the procedure. The Precision Descent eliminates one clearance used in the earlier study, and greatly reduces the length of the remaining clearances. This was accomplished by using non-standard clearance phraseology that relies on a published procedure chart for correct interpretation. Eight type-rated flight crews flew eight Precision Descents in a Boeing 747-400 simulator. No training was provided: crews received either a procedure chart or a procedure chart with a flight manual bulletin describing procedure techniques. Video and digital data were recorded for each descent. Preliminary results indicate that moving information from the verbal clearance to the chart was successful: the shorter clearances and the procedure's default crossing restriction were understood, and procedure compliance of 81% was observed across all descents on six independent compliance measures. Some readback errors and procedure misunderstandings were still observed, however; particularly when descent clearances deviated from the normative procedure.