Synopsis On takeoff from FRA, the A320 encountered turbulence which tipped the left wing down. The captain, who was the pilot flying, corrected with right stick, whereupon the aircraft banked even further left. The left wingtip was said to have come within two meters of hitting the ground. The first officer saw what was going on, cut out the captain's stick inputs (there is a "takeover" button), and took over control. The aircraft landed again at FRA, after the crew checked out the first officer's control at altitude.
The incident was due to the miswiring of the captain's control stick during maintenance. Left stick normally commands left bank; right stick right bank. In this case, left stick commanded right bank and right stick left bank. The captain's control inputs to the turbulence-induced left bank thus made the left bank steeper, rather than correcting it.
Why did maintenance make such a mistake? Because they misread the document trail showing them how the wiring was to be connected on this particular aircraft. When they returned the aircraft to the flight line from maintenance, they tested the controls from the copilot's side, but not from the captain's side, although they had rewired his sidestick to the flight control computers. And the flight crew did not detect the anomaly during pre-flight check of the controls.
Reversing the control connections happens every so often to conventional aircraft, in particular to general aviation aircraft. Some pilots cope; some pilots don't. Best to make sure you know about it *before* takeoff. When I reclaimed my Piper Archer after maintenance, I always checked to make sure the control surfaces moved in the right direction. People thought that this kind of thing could not happen to fly-by-wire aircraft. Surprise: it can and did.
I wrote about this incident in the Risks Forum, in articles A320 Incident in Risks 21.48, before the Final Report had been published, and A320 Incident in Risks 23.24, afterwards. As I said in my Risks 23.24 note, the explanation still leaves me puzzled. I still do not understand why the two ELAC flight control computers did not receive contradictory bank inputs from the captain's sidestick control (only one plug, to one ELAC, was rewired) and thus report an anomaly at preflight check or before.