FDAI logo   ::  Site Map  ::   
Home  |  About This Website  |  Contact Us
About This Website » ASRS Incident Report Analysis » Incidents related to Issue

Incidents related to Issue 2 incidents related to this issue.

Issue: automation level decisions may be difficult (Issue #103)
Description: It may be difficult for pilots to decide what levels of automation are appropriate in specific circumstances, possibly increasing pilot workload.

  1.  
  2. Accession Number: 130037
    Synopsis: ALT DEVIATION. DESCENT CROSSING RESTRICTION NOT MET.
    Narrative: WE WERE FLYING FROM LAS TO SNA IN AN MLG EQUIPPED WITH A SINGLE FMS CDU ON THE LEFT SIDE OF THE FORWARD ELECTRONIC PANEL (FORWARD OF THE THRUST LEVERS), AND ELECTRO-MECHANICAL FLT INSTRUMENT DISPLAYS. WE HAD BEEN CLEARED TO FLY THE KAYOH ONE ARR, AND HAD PROGRAMMED THE FMC FOR A VNAV DSNT TO CROSS DAWNA (HDF 353/29) AT 13,000, AND THEN CROSS KAYOH (SLI 075/23) AT 210 KTS AND 8000. BEFORE REACHING DAWNA WE WERE CLEARED DIRECT TO HDF. ZLA BROADCAST A SIGMET THAT FORECAST OCCASIONAL SEVERE TURBULENCE BELOW 10,000 IN THE AREA THAT WE WOULD BE MAKING OUR FINAL DSNT, FOLLOWED IMMEDIATELY BY A CLRNC FOR US TO CROSS 20 NE OF HDF AT 14,000. WE BOTH REMARKED THAT HDF WAS NOT SHOWING A DME SIGNAL, SO RAW DATA DISTANCE WAS NOT AVAILABLE. I ASKED THE F/O IF HE WOULD LIKE ME TO ENTER THE RESTRICTION INTO THE FMC (SINCE IT'S AWKWARD TO MAKE DATA ENTRIES FROM HIS SIDE OF THE COCKPIT) AND ENTERED HDF 010/20 AT 14,000. THE F/O INDICATED AGREEMENT WITH THE DATA ENTRY, AND I TOLD HIM I'D BE "OFF THE RADIO" WHILE I DISCUSSED THE ANTICIPATED TURBULENCE WITH THE CABIN CREW. DURING THAT DISCUSSION I OBSERVED THE F/O INITIATE A RAPID DSNT, CHECKED THE CDU, AND WAS SURPRISED TO SEE THAT THE WAYPOINT I HAD JUST ENTERED WAS NO LONGER DISPLAYED (SINCE WE HAD JUST PASSED IT). THE F/O TOLD ME THAT WE HAD CROSSED THE 20 MI FIX PASSING THROUGH 17,000, AND THAT CENTER'S ONLY COMMENT WAS, "YOU MISSED IT". I BELIEVE THAT THIS INCIDENT WAS NOT DUE TO ANY INATTENTION, BUT RATHER TO HUMAN FACTORS PROBLEMS INVOLVED WITH THE AUTOMATED COCKPIT AND ATC. I ALSO BELIEVE THAT IT WOULDN'T HAVE HAPPENED IN AN OLDER-TECHNOLOGY ACFT (WHICH I ALSO FLY). FOR EXAMPLE: 1) IN THE OLDER-TECHNOLOGY ACFT, I WOULDN'T'T HAVE ACCEPTED THIS CLRNC, ONCE I REALIZED THAT HDF HAD NO DME. I ASSUME THAT THE CTLR WOULDN'T HAVE ISSUED IT UNLESS HE KNEW THE ACFT WAS RNAV EQUIPPED. 2) I'M NOT USED TO RECEIVING CLRNCS TO A "DISTANCE" POINT ASSOCIATED WITH A NON-DME FAC. I ASSOCIATE THE PHRASEOLOGY, "CROSS TWENTY NE OF XYZ" WITH A DME FIX, WHERE I CAN QUICKLY LOOK AT A RAW-DATA DME DISPLAY AND MAKE AN IMMEDIATE JUDGEMENT THAT A DSNT MUST BE INITIATED. I LOOKED FOR RAW DATA. IT WASN'T AVAILABLE, SO I MENTALLY SHIFTED FROM THE OLD-TECHNOLOGY (RAW DATA) MODE TO THE AUTOMATED COCKPIT (DATA ENTRY AND COMPUTED SOLUTION) MODE. 3) MOST CLRNCS I'VE RECEIVED THAT REQUIRE RNAV INVOLVE NAMED FIXES WHICH ARE USUALLY IN THE FLT PLAN OR DATABASE. I CAN MAKE A RAPID (4 KEYSTROKES) DATA ENTRY FOR A RNAV POSITION/ALT IF IT WAS PREVIOUSLY PROGRAMMED INTO THE ROUTE, IE, DIRECT HDF. THE CLRNC IN THIS INCIDENT REQUIRED 15 KEYSTROKES, AND TOOK CONSIDERABLY MORE TIME. 4) WHEN I FLY AN AUTOMATED ACFT, I HAVE TO SPEND SOME AMOUNT OF TIME PROGRAMMING AND MONITORING THE COMPUTER. THE DECISION OF WHEN TO DO THIS AND WHEN IT ISN'T APPROPRIATE IS A SUBTLE ONE, AND LENDS ITSELF TO GENERALITIES SUCH AS, "DON'T PROGRAM IN THE TERMINAL AREA." IN THIS CASE, "DON'T PROGRAM IN SOUTHERN CALIFORNIA" WOULD HAVE BEEN APPROPRIATE. THE DATA WE NEEDED (PRESENT DISTANCE TO HDF) WAS AVAILABLE IN THE CDU, SO IF WE HAD TREATED IT AS IF WE DIDN'T HAVE RNAV AVAILABLE WE MIGHT HAVE BEEN ABLE TO MAKE THE MENTAL CALCULATION IN TIME TO MEET THE RESTRICTION. IN THE ENRTE ENVIRONMENT HOWEVER, EXPERIENCE LED ME TO THE SUBCONSCIOUS DECISION THAT THERE WAS TIME AVAILABLE TO REPROGRAM THE FMC. 5) THE CTLR ISSUED A CLRNC THAT WAS PROBABLY DELAYED DUE TO THE SIGMET BROADCAST. AS IT WAS, THE CLRNC WAS MADE SO CLOSE TO THE FIX THAT BY THE TIME IT WAS PROGRAMMED INTO THE FMC AND DSNT COMPUTATIONS MADE, IT WAS NOT POSSIBLE TO MEET THE RESTRICTION. 6) IT WASN'T UNTIL THE NEXT DAY AND A DISCUSSION EXPLORING BOTH OF OUR THOUGHT PROCESSES THAT WE FOUND ANYTHING WE (AS OPPOSED TO THE CTLR) COULD HAVE DONE DIFFERENTLY UNDER THE CIRCUMSTANCES TO PREVENT THIS INCIDENT. WHEN EITHER OF US IS THE PLT FLYING AND MAKES A DATA ENTRY, WE EACH MAKE A MENTAL VALIDITY CHECK USING SOME RULE OF THUMB, IE, "I'M 13 MI FROM THE FIX AND 6000' HIGH, SO I'VE GOT TO START DOWN NOW!" IN THIS CASE, ALTHOUGH I MADE THE DATA ENTRY I DIDN'T MAKE THE VALIDITY CHECK BECAUSE I WAS NOT FLYING, AND PREOCCUPIED WITH MY CONCERN FOR THE SAFETY OF THE PAX AND CABIN CREW IN THE CONTEXT OF THE EXPECTED SEVERE TURBULENCE. I HAD SUBCONSCIOUSLY PLACED A HIGHER PRIORITY ON COMMUNICATING WITH THE CABIN CREW THAN ON DOUBLE-CHECKING THE COMPUTER. IT HAS NOT ESCAPED ME THAT NOT DOUBLE-CHECKING THE COMPUTER OR THE PLT FLYING CAN LEAD TO A BUSTED ALT AND HENCE A MIDAIR COLLISION, WHICH SHOULD HAVE BEEN A HIGHER PRIORITY. THE F/O INDICATED THAT SINCE I HAD MADE THE DATA ENTRY HE ASSUMED THAT THE DATA WAS VALID (IT WAS, BUT TOO LATE) AND DIDN'T DOUBLE-CHECK IT.

  3.  
  4. Accession Number: 396329
    Synopsis: A B767 CREW EXPERIENCES A MAP SHIFT WHILE ATTEMPTING TO ALIGN WITH THE RWY 25R ILS LOC. ACFT HAD BEEN ON THE APCH FOR RWY 24R AND HAD BEEN SHIFTED OVER TO THE S COMPLEX BY APCH CTLR.
    Narrative: THE REMARKS SECTION OF OUR FLT PLAN CONTAINED A WARNING TO WATCH FOR AN FMS MAP SHIFT INBOUND TO LAX. IT SAID SEVERAL ACFT HAD EXPERIENCED MAP SHIFTS DURING THE ARR. WE WERE CLRED DIRECT 'CIVET' INTXN AND WERE EXPECTING THE 'CIVET' ARR TO RWY 25L. THE FMS WAS PROGRAMMED FOR A 'CIVET' ARR AND ILS RWY 25L. ATC THEN CHANGED OUR CLRNC TO THE 'MITTS' ARR, DIRECT 'SNRKL' INTXN, CROSS 'SNRKL' AT OR ABOVE 12000 FT, CLRED ILS RWY 24R. WE RETUNED THE NAVAIDS AND REPROGRAMMED THE FMS FOR A 'MITTS' ARR TO RWY 24R. WE WERE ISSUED TFC TO OUR L SEVERAL TIMES BUT NEVER DID SEE THE TFC. ATC THEN OFFERED, AND WE ACCEPTED RWY 25R. I TUNED THE LOC AND WENT TO A RAW DATA DISPLAY ON MY HSI, WHILE THE CAPT LOOKED FOR THE RWY 25R APCH PLATE AND THEN SELECTED THE ILS RWY 25R IN THE FMS. WE WERE INSTRUCTED TO JOIN THE RWY 25R LOC, CROSS 'MUSIK' INTXN AT OR ABOVE 8000 FT, CLRED ILS RWY 25R. I DON'T RECALL SEEING OUR POS IN RELATION TO THE RAW DATA DISPLAY OF THE LOC BECAUSE I WAS DISTRACTED AT THIS POINT: THE ACFT MADE AN ABRUPT L TURN, AND THEN WE GOT A TCASII RA TO 'REDUCE DSCNT.' I BELIEVE THE ABRUPT L TURN WAS CAUSED BY A MAP SHIFT THAT BECAME APPARENT A BIT LATER. ANYWAY, WE RESPONDED TO THE TCASII WARNING. THEN I SAW THE ARPT VISUALLY AND REALIZED WE WERE WELL S OF THE LOC WHEN I EXPECTED TO BE N OF THE LOC. WE INITIATED A R TURN BACK TOWARD THE LOC IMMEDIATELY, AND INFORMED ATC THAT WE WERE CORRECTING. AT THIS POINT I COULD SEE THAT THE CAPT'S DISPLAY SHOWED US ABOUT 5 MI N OF THE LOC, WHEN IN REALITY WE WERE S OF THE LOC. THE ARR AND RWY CHANGES MADE DETECTING THE MAP SHIFT ALMOST IMPOSSIBLE UNTIL IT WAS TOO LATE. I THINK THE PNF SHOULD MONITOR RAW DATA AT ALL TIMES DURING ARRS TO LAX ARPT. SUPPLEMENTAL INFO FROM ACN 396417: MOMENTS PASSED AS WE SET UP FOR THAT CLRNC. ATC CONTINUED TO CALL WITH TFC, AIRSPD RESTRS, AND FREQ CHANGES, REQUIRING RESPONSES AND INTERRUPTIONS TO OUR FMS SET-UP. WE WERE GIVEN ANOTHER FREQ CHANGE. SOCAL APCH CTL IMMEDIATELY OFFERED US RWY 25R. ALTHOUGH ALMOST REFUSING THE OFFER, IT WAS ACCEPTED FOR THE SHORTER TAXI TIME TO THE GATE. (ALMOST REFUSED DUE TO MULTIPLE CLRNC CHANGES ALREADY AND DIFFICULTY, DUE TO TIME CONSTRAINTS AND UNFAMILIARITY, IN NEW SETUPS FOR EACH CHANGE AND MULTIPLE CALLS FROM ATC INHIBITING OUR CLRNC PROCESSING.) SOCAL APCH THEN CLRED US TO INTERCEPT THE RWY 25R LOC. LOOKING AT MY HSI MAP, I EXPECTED TO HAVE TO TURN L SLIGHTLY TO INTERCEPT THE LOC. THE AUTOPLT ALREADY WAS/HAD BEEN IN A LOW-BANK TURN TO THE L. THIS PUZZLED ME FOR A MOMENT. THE INTERCEPT ANGLE WAS ALREADY TOO GREAT SO I INTERCEDED AND REDUCED THE ANGLE. SOCAL APCH HAD CLRED US FOR THE APCH. ALMOST IMMEDIATELY OUT OF 8000 FT, WE HAD A TCASII ALERT, 'REDUCE DSCNT.' WHEN I LOOKED AT THE TCASII, THE ALERT SOUNDED ONE MORE TIME AS I SAW A TARGET PASS 500 FT, MORE OR LESS UNDER US, ALSO IN A DSCNT. AT THAT TIME A PLT JUMP SEAT RIDER BEHIND ME POINTED OUT BY VISUAL REF TO THE ARPT THAT WE WERE S OF BOTH RWY 25L AND RWY 25R LOCS AND GOING FURTHER S. I IMMEDIATELY WENT TO RAW DATA, MADE A CORRECTION BACK TO THE R TO INTERCEPT RWY 25R LOC AND FURTHER REDUCED OUR RATE OF DSCNT. AUTOPLT LOC WAS ALSO ARMED. WITHIN SECONDS, WE WERE ON THE LOC AND THE REMAINDER OF THE FLT PROGRESSED NORMALLY. NEEDLESS TO SAY, WE SPENT SOME TIME ANALYZING THIS. OUR FIRST REALIZATION WAS THAT WE'D EXPERIENCED A 'MAP SHIFT,' WITHOUT RECOGNIZING IT IMMEDIATELY. UPON REALIZING THAT WE WERE CHASING A PHANTOM LOC TO THE S (L), WE RECOGNIZED THE SHIFT. MINS LATER, AT THE GATE, MAINT CONFIRMED A SHIFT OF AT LEAST 2 1/2 MI. ON REFLECTION, WE MIGHT HAVE SAVED SOME OF THIS IRREGULAR OP HAD WE: 1) ARMED THE AUTOPLT FOR LOC INTERCEPTION IMMEDIATELY AFTER RECEIVING THE CLRNC AND DIALING IN THE LOC FREQ. 2) DOWNLOADED SOME OF THE AUTOMATION RIGHT AWAY AND GONE TO MANUAL AND RAW DATA MODE. 3) REFUSED THE OFFER OF RWY 25R SINCE WE WERE SET UP FOR RWY 24R. NORMALLY, WE DON'T INTERCEPT THE LOC TOO FAR OUT BECAUSE IT'S TOO UNSTABLE AND CAUSES THE ACFT TO WING WALK TOO MUCH. INSTEAD, WE USE THE FMS CALCULATED LOC, AND ARM THE LOC WHEN IN CLOSER. STANDARD PROC ALSO CALLS FOR 1 PLT TO BE ON RAW DATA FOR LOC INTERCEPT. THAT WAS DONE, BUT WE WERE PLAYING CATCH-UP WITH THE CLRNC AND FAILED TO CATCH THE RAW DATA DISCREPANCY RIGHT AWAY. ALL OF THIS HAPPENED IN A VERY SHORT TIME FRAME, COMPRESSED AND COMPLICATED BY AN UNUSUALLY HIGH NUMBER OF CALLS FROM ATC. ALTHOUGH RAW DATA WAS AVAILABLE TO CORRECT US, WE DID NOT SEE IT IN THE FEW SECONDS AVAILABLE BEFORE WE REALIZED IT ANYWAY. THE MAP SHIFT WAS A SERIOUS COMPLICATION. IT HAD ALREADY POINTED THE ACFT TOWARDS THE SW BEFORE WE SAW IT. WITHOUT ALL THE CLRNC CHANGES, THE MAP SHIFT MORE THAN LIKELY WOULD HAVE BEEN CAUGHT MUCH MORE QUICKLY, IF NOT IMMEDIATELY. WITH PAST KNOWN MAP SHIFTS, I WOULD RECOMMEND THAT 1 PLT MONITOR THE ENTIRE ARR AND APCH, AS WE DO IN MEXICO CITY. NEVER LET ATC FURTHER COMPLICATE OR RUSH YOUR ARR/APCH. AND DOWNLOAD TO THE APPROPRIATE LEVEL OF AUTOMATION IMMEDIATELY WHEN NECESSARY. WITH MY 11 YRS EXPERIENCE IN GLASS COCKPITS, I REALLY THOUGHT THAT I WAS MORE IMMUNE TO THIS KIND OF THING HAPPENING TO ME THAN MOST. LET ME TELL YOU. CIRCUMSTANCES CAN CONSPIRE TO TRAP ANY UNSUSPECTING VICTIM.
Flight Deck Automation Issues Website  
© 1997-2013 Research Integrations, Inc.