Narrative:

Both of us were somewhat fatigued due to circumstances prior to the trip. We had just switched to toronto arrival control on the youth 2 arrival and were originally planning on a runway 6R approach based on recent experience; even though ATIS was advertising runway 5. Both of us agreed to leave runway 6R programmed until we had a final determination on the runway assignment from the arrival controller. Immediately after we switched to arrival; runway 5 was assigned and we began setting up for that approach. This occurred about halfway between the youth and verko intxns. When I entered the runway 5 approach into the mcdu; the aircraft began a right turn as if to return all the way to buf and re-begin the arrival procedure. At this point; both of us were heads down setting up the approach and didn't notice the undesired turn until queried by the controller. I began an immediate left turn back toward the intended arrival route and we continued on for the runway 5 approach with radar vectors. Neglected to verify inputs to the FMC for the new runway (might have remembered that the aircraft would want to return to buf); should have put the autoplt in a different lateral mode to keep it on the arrival route. No one was watching the aircraft...both of us were reviewing the approach. Make certain that one of us is watching the aircraft during the runway change inputs. In my defense; while this navigation system's quirk of wanting to go all the way back to the beginning of an arrival procedure because of a runway change is well known; it is still a very illogical and counterintuitive one...and potentially dangerous! Can't someone talk some sense into the people that design these systems about this? I needed a new runway; nothing more! The arrival for either of the runways in question here is exactly the same and ultimately ends in a vectored approach anyway. I never wanted to leave the arrival route we were already on. I could have set the autoplt to heading and prevented the erroneous turn; but we were also only a few miles from a 90 degree left turn on the arrival route; which could very easily also have led to a navigation error had the heading not been manually changed at that point. I realize that the primary cause here was not watching the aircraft; but this design is a built-in trap. At best; this system needlessly adds to my cockpit workload; when its apparent objective is supposedly to reduce it. At worst; it can lead to a navigational error...as it did here! And I know I'm not the first with this.

Related reports:

Google
 

Original NASA ASRS Text

Title: A FATIGUED E190 PLT RPTS A TRACK DEV AFTER THE FMC ATTEMPTED TO FLY THE ACFT BACK TO THE STAR'S BEGINNING WHILE A CLOSE IN RWY CHANGE WAS MADE.

Narrative: BOTH OF US WERE SOMEWHAT FATIGUED DUE TO CIRCUMSTANCES PRIOR TO THE TRIP. WE HAD JUST SWITCHED TO TORONTO ARR CTL ON THE YOUTH 2 ARR AND WERE ORIGINALLY PLANNING ON A RWY 6R APCH BASED ON RECENT EXPERIENCE; EVEN THOUGH ATIS WAS ADVERTISING RWY 5. BOTH OF US AGREED TO LEAVE RWY 6R PROGRAMMED UNTIL WE HAD A FINAL DETERMINATION ON THE RWY ASSIGNMENT FROM THE ARR CTLR. IMMEDIATELY AFTER WE SWITCHED TO ARR; RWY 5 WAS ASSIGNED AND WE BEGAN SETTING UP FOR THAT APCH. THIS OCCURRED ABOUT HALFWAY BETWEEN THE YOUTH AND VERKO INTXNS. WHEN I ENTERED THE RWY 5 APCH INTO THE MCDU; THE ACFT BEGAN A R TURN AS IF TO RETURN ALL THE WAY TO BUF AND RE-BEGIN THE ARR PROC. AT THIS POINT; BOTH OF US WERE HEADS DOWN SETTING UP THE APCH AND DIDN'T NOTICE THE UNDESIRED TURN UNTIL QUERIED BY THE CTLR. I BEGAN AN IMMEDIATE L TURN BACK TOWARD THE INTENDED ARR RTE AND WE CONTINUED ON FOR THE RWY 5 APCH WITH RADAR VECTORS. NEGLECTED TO VERIFY INPUTS TO THE FMC FOR THE NEW RWY (MIGHT HAVE REMEMBERED THAT THE ACFT WOULD WANT TO RETURN TO BUF); SHOULD HAVE PUT THE AUTOPLT IN A DIFFERENT LATERAL MODE TO KEEP IT ON THE ARR RTE. NO ONE WAS WATCHING THE ACFT...BOTH OF US WERE REVIEWING THE APCH. MAKE CERTAIN THAT ONE OF US IS WATCHING THE ACFT DURING THE RWY CHANGE INPUTS. IN MY DEFENSE; WHILE THIS NAV SYSTEM'S QUIRK OF WANTING TO GO ALL THE WAY BACK TO THE BEGINNING OF AN ARR PROC BECAUSE OF A RWY CHANGE IS WELL KNOWN; IT IS STILL A VERY ILLOGICAL AND COUNTERINTUITIVE ONE...AND POTENTIALLY DANGEROUS! CAN'T SOMEONE TALK SOME SENSE INTO THE PEOPLE THAT DESIGN THESE SYSTEMS ABOUT THIS? I NEEDED A NEW RWY; NOTHING MORE! THE ARR FOR EITHER OF THE RWYS IN QUESTION HERE IS EXACTLY THE SAME AND ULTIMATELY ENDS IN A VECTORED APCH ANYWAY. I NEVER WANTED TO LEAVE THE ARR RTE WE WERE ALREADY ON. I COULD HAVE SET THE AUTOPLT TO HEADING AND PREVENTED THE ERRONEOUS TURN; BUT WE WERE ALSO ONLY A FEW MILES FROM A 90 DEG L TURN ON THE ARR RTE; WHICH COULD VERY EASILY ALSO HAVE LED TO A NAV ERROR HAD THE HEADING NOT BEEN MANUALLY CHANGED AT THAT POINT. I REALIZE THAT THE PRIMARY CAUSE HERE WAS NOT WATCHING THE ACFT; BUT THIS DESIGN IS A BUILT-IN TRAP. AT BEST; THIS SYSTEM NEEDLESSLY ADDS TO MY COCKPIT WORKLOAD; WHEN ITS APPARENT OBJECTIVE IS SUPPOSEDLY TO REDUCE IT. AT WORST; IT CAN LEAD TO A NAVIGATIONAL ERROR...AS IT DID HERE! AND I KNOW I'M NOT THE FIRST WITH THIS.

Data retrieved from NASA's ASRS site as of May 2009 and automatically converted to unabbreviated mixed upper/lowercase text. This report is for informational purposes with no guarantee of accuracy. See NASA's ASRS site for official report.