Narrative:

I would like to discuss a significant safety issue that is incorrectly described to the pilots. It is 'alt hold' engagement during the takeoff roll. Before I go into this discussion; please allow me to state my background. I hold a B.south. In aerospace engineering; I am a senior officer in the military and I held jobs as safety department head; senior iut evaluator and an equivalent of assistant model manager. Over last few years I observed 4 instances where 'alt hold' was engaged during the takeoff run; 3 in foreign airports and 1 domestically. The domestic event; I actually saw it change at about 80 KTS from VNAV to altitude. Before each flight; I brief the crew [on] what [I've] seen and the events that occur during this event. This is my brief: during the pre-departure brief; we as crew verify that we have everything setup properly. This is done prior to push from the gate and includes: vertical and lateral modes; usually in LNAV and VNAV; altitude is set to our clearance altitude; flight director's (FD's) are indicating properly; takeoff/go-around with LNAV/VNAV below and verify all other items indicate properly. During the takeoff roll; about 60-80 KTS; the VNAV changes to altitude; altitude hold button lights up on mode control panel (MCP) and FD's are providing incorrect guidance. As the aircraft takes off; the FD's still provide incorrect guidance (fly down to takeoff altitude) until passing 300-400 ft. At that time the indication and FD guidance changes to VNAV speed giving proper indication. One can fly through the improper indication or turn off then on both flight directors for proper guidance. If autopilot is turned on at 200 ft the aircraft will descend to the takeoff altitude and impact the ground. Worst-case scenario is mumbai; maximum gross weight; at night; in monsoon and engine failure at V1. We are instructed to turn autopilot 'on' at 200 ft. A lot of things are happening at that time and my recommendation is to verify everything and not turning autopilot on until passing 500 ft. Many pilots that I brief stated that they been told that the initial altitude was set to zero in MCP. I can attest that is not the case. In mumbai; the aircraft is warm and when entering the cockpit the MCP altitude is usually set at missed approach altitude and not zero. Discussing this issue with the foqa personnel; I had been told that the foqa was registering zero during the engine starts. I can attest that it is not the case. In my 25-year flying career I have yet to see initial altitude set to zero. It might be set zero at the gate; but after the briefs it is always set to initial clearance altitude. That is not what foqa data is indicating and not what is passed to the pilot group. What I see are two separate problems. One: during the takeoff run; VNAV on its own; transitions to altitude (hold) and gives improper guidance. Two: foqa is recording improper indication from MCP. My recommendation is to brief the pilots that this is happening during the takeoff run with proper settings prior to taking the runway and not to engage the autopilot until 500 ft after verifying that all indications are correct. In addition to the 'alt' (hold) problem; foqa data recorder should be investigated why it is recording incorrect data. I see them as two completely separate issues.

Google
 

Original NASA ASRS Text

Title: A B777 pilot reported the Mode Control Panel (MCP) made an uncommanded mode change from VNAV to ALT HOLD during the takeoff roll at about 80 KTS.

Narrative: I would like to discuss a significant safety issue that is incorrectly described to the pilots. It is 'Alt Hold' engagement during the takeoff roll. Before I go into this discussion; please allow me to state my background. I hold a B.S. in Aerospace Engineering; I am a Senior Officer in the military and I held jobs as Safety Department Head; Senior IUT Evaluator and an equivalent of Assistant Model Manager. Over last few years I observed 4 instances where 'Alt Hold' was engaged during the takeoff run; 3 in foreign airports and 1 domestically. The domestic event; I actually saw it change at about 80 KTS from VNAV to ALT. Before each flight; I brief the crew [on] what [I've] seen and the events that occur during this event. This is my brief: During the pre-departure brief; we as crew verify that we have everything setup properly. This is done prior to push from the gate and includes: vertical and lateral modes; usually in LNAV and VNAV; altitude is set to our clearance altitude; Flight Director's (FD's) are indicating properly; takeoff/go-around with LNAV/VNAV below and verify all other items indicate properly. During the takeoff roll; about 60-80 KTS; the VNAV changes to ALT; ALT HOLD button lights up on Mode Control Panel (MCP) and FD's are providing incorrect guidance. As the aircraft takes off; the FD's still provide incorrect guidance (fly down to takeoff altitude) until passing 300-400 FT. At that time the indication and FD guidance changes to VNAV SPEED giving proper indication. One can fly through the improper indication or turn OFF then ON both flight directors for proper guidance. If autopilot is turned on at 200 FT the aircraft will descend to the takeoff altitude and impact the ground. Worst-case scenario is Mumbai; maximum gross weight; at night; in Monsoon and engine failure at V1. We are instructed to turn autopilot 'ON' at 200 FT. A lot of things are happening at that time and my recommendation is to verify everything and not turning autopilot on until passing 500 FT. Many pilots that I brief stated that they been told that the initial altitude was set to zero in MCP. I can attest that is not the case. In Mumbai; the aircraft is warm and when entering the cockpit the MCP altitude is usually set at missed approach altitude and not zero. Discussing this issue with the FOQA personnel; I had been told that the FOQA was registering zero during the engine starts. I can attest that it is not the case. In my 25-year flying career I have yet to see initial altitude set to zero. It might be set zero at the gate; but after the briefs it is always set to initial clearance altitude. That is not what FOQA data is indicating and not what is passed to the pilot group. What I see are two separate problems. One: During the takeoff run; VNAV on its own; transitions to ALT (Hold) and gives improper guidance. Two: FOQA is recording improper indication from MCP. My recommendation is to brief the pilots that this is happening during the takeoff run with proper settings prior to taking the runway and not to engage the autopilot until 500 FT after verifying that all indications are correct. In addition to the 'Alt' (hold) problem; FOQA data recorder should be investigated why it is recording incorrect data. I see them as two completely separate issues.

Data retrieved from NASA's ASRS site as of July 2013 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.