Readback · ROT-01 · Spatial orientation
A cube carries a mark on one face. You hear a run of ninety-degree turns — up, down, left, right — and must say where the mark ends up. It is the ROT subtest of the DLR battery, and the same ability sits behind the Aon cut-e spatial orientation modules. Nothing is written down and the cube is not shown while it turns.
The cube is shown once at the start so you can see where the mark is. After that it disappears and the turns are only spoken — exactly as in the real subtest, where you are given the moves and must hold the object in your head.
| # | Start | Sequence | Yours | Correct |
|---|
The cube rotation test — usually written ROT — is a spatial orientation subtest in the DLR battery run by the German Aerospace Center. Reported users of the DLR route include Lufthansa and Lufthansa City Airlines, Austrian Airlines, Air Dolomiti, Luxair, ADAC Air Rescue and the Swiss Air Force.
Aon cut-e measures the same underlying ability through different instruments — its spatial orientation modules use a gyro-style attitude display and a relative bearing indicator rather than a cube. If you are preparing for cut-e rather than DLR, the interface will look nothing like this page, but the skill you are training is the same one.
What the subtest actually measures is not cleverness with cubes. It is whether you can hold a spatial model in your head while new information keeps arriving and overwriting it — the same demand a pilot faces holding a mental picture of the aircraft's attitude while reading, listening and deciding. That is why the test is spoken and paced rather than printed and untimed.
Almost everyone starts by trying to picture the whole cube turning. That is the slow way and it collapses after three or four moves. Track one thing: the direction the mark is facing.
Up and down move the mark through Front → Top → Back → Bottom and round again. Left and right move it through Front → Right → Back → Left. Down and left simply run those cycles backwards. Four moves in one direction always returns you to where you started.
A face lying on the axis of the turn stays put. On a left or right turn, Top and Bottom never move. On an up or down turn, Left and Right never move. Roughly a third of moves in a random run change nothing at all — recognising them instantly is free speed.
An up followed by a down cancels, as does left then right. In a long run you can often collapse the sequence to two or three effective moves before you finish hearing it. Listen for the cancellation rather than dutifully applying every step.
Holding a picture is expensive; holding a word is cheap. Repeat the current face to yourself as a single word — "top… back… left" — and update it on each command. Verbal encoding survives interference far better than a visual one.
Start at four seconds per move with three moves. When that is comfortable, add moves before you shorten the gap — length taxes memory, speed taxes retrieval, and memory is the one that fails first.
If you lose the thread mid-run, take the last face you were sure of and apply whatever you remember from there. A reasoned guess scores; a blank does not, and stopping to recover the whole chain costs you the next question too.