Autopilots Don’t Always Work as Advertised

Sometimes procedures don’t work out the way they are supposed to. An over-enthusiastic autopilot served as a catalyst for one such realization. Namely, that turning the autopilot on does not relieve a pilot from the task of flying the airplane. This morning’s flight served as clear reminder of that fact.


It was my leg to fly an early morning departure from Fort Walton Beach, Florida to Memphis, Tennessee. The airplane was in good working order, except for the windshield wipers which had been deferred and placarded inoperative. Since the weather was forecast to be “clear and a million” all the way, this was a non-issue.

Restricted Airspace Near Ft. Walton

Restricted Airspace Near Ft. Walton

Although our company profiles allow (and some would say encourage) the autopilot to be engaged at 600 feet above ground level, I normally hand fly the airplane to 10,000 feet before I consider engaging the autopilot, but today was different. Due to a high level of naval activity, restricted airspace speckles the Florida Panhandle. Furthermore, the proximity of Destin, a popular GA destination makes for a number of VFR targets that flight crews must see and avoid. In the interest of safety, I opted to reduce my workload by allowing the autopilot to do the flying bit while I focused on dodging restricted airspace and rogue Cessnas.

That was the plan anyway. Our takeoff profile requires that we rotate at Vr, pitch up 10º, then maintain an airspeed of V2+10 for the first 1000 feet. In this case, Vr was 132 knots with V2 at 139, which would require a climb speed of 149 knots.


Passing 600 feet, I called “autopilot on,” and the captain engaged the autopilot. By now, the airplane was at 160 knots, and the autopilot should have smoothly pitched up (or down) to maintain 160 knots in the climb. Much to my surprise, the aircraft pitched up aggressively, resulting in a rapid loss of airspeed as indicated by a growing trend vector on my primary flight display.

As the speed fell below 160 knots, the autopilot began to slowly pitch down. Unfortunately, the pitch moment was too slow, prompting me to disconnect the autopilot as the speed dipped below 149 knots. By the time the airspeed began to increase, we were as slow as 140 knots, just one knot above V2!

The lesson to take from this story is that the autopilot does not always behave appropriately. It is the pilot’s responsibility to monitor the situation and take corrective action if necessary. Shortly after engaging the autopilot, keep your hand on the control wheel and be ready to disconnect the autopilot and fly the airplane at a moment’s notice.

Sometimes you just have to roll the hard six. Keep your eyes open and fly safely.


Related Posts:

Tags: ,
fold-left fold-right
About the author
Pat Flannigan is a professional pilot and aviation blogger. He has been flying for fifteen years and is currently working as an airline pilot in the United States.

4 Replies to Autopilots Don’t Always Work as Advertised

  1. treetopflyer says:

    never trust autopilot….I prefer to fly it, by hand, as far into the crash as possible.

  2. Anything can go hectic at any time so it’s good to monitor everything closely. Just look at the Turkish 737 crash… faulty altimeter/auto-pilot signal.

Please, share your thoughts and opinions

%d bloggers like this: