CubePick reach, layout 4

cubepick-reach-layout-4cubepickPoint end effector in the unit squareManipulationeasy

Reference solution, 8 of 80 steps.

Instruction

A point end effector moves in the unit square (x from 0 to 1, y from 0 to 1). It starts at (0.1, 0.1); a cube lies somewhere in the far quadrant (both coordinates between 0.6 and 0.9). This is inspect-robots' built-in cubepick-reach benchmark, layout 4.

Task

Reach the cube.

Success: the episode is scored as solved the first time the end effector is within 0.05 of the cube (upstream's success_at_end with termination on success). Calling robo done earlier ends the episode unsolved.

Controls. robo act DX DY moves the end effector by (DX, DY), each clipped to +-0.1; positions stay inside the unit square. --repeat N applies the same move N times.

Observation. robo observe reports eef_pos, cube_pos and their distance. robo observe --image saves the top camera: the cube is the green square, the end effector the red one; +x points right and +y points down in the image.

The step budget is 80 steps.

How the robot is controlled and scored

You are controlling a simulated robot. Read the task below, then solve it by running the robo command in your shell (start with robo info and robo observe). Keep going until the task is done, then call robo done once. Do not stop to ask questions; there is no human to answer.

How to control the robot

You are the robot's policy. You act only through the robo command in your shell. There is no other way to move the robot, and you cannot read or change the simulator, the scoring, or other files to succeed; the episode server judges the final physical state itself.

robo info                          # the robot, its sensors, action groups, skills and step budget
robo observe                       # robot and scene state as numbers
robo observe --image [--camera C]  # also saves a camera image and prints its path (open it to look)
robo act V1 V2 ... [--repeat N]    # one low-level action (the action groups under Controls), applied N times (N <= 50)
robo skill NAME ARG ...            # run a skill listed by `robo info`; it runs until it finishes and reports the result
robo done "short summary"          # end the episode and ask for scoring
robo give-up "reason"              # end the episode without claiming success

Positions are in metres in the world frame (+z up); angles are in degrees unless a field says otherwise.
The episode has a fixed step budget (see robo info); every simulated control step counts, including the steps a skill runs.
Skills are ordinary controllers: they can fail, stop early or be blocked by the scene. Read what they report and re-observe.
Success is judged about 10 steps after you call robo done, with the robot holding still (each action group's hold value: zero for velocity and delta commands, full brake for a car), so the goal must still be true when the robot stops.
Call robo done exactly once when finished.

Run this task

ROBOUSE_ORACLE_TOKEN=$(openssl rand -hex 16) \
  bench eval run \
    -d robocurve/cubepick-reach@0.2 \
    --registry https://robouse.ai/hub/registry.json \
    --agent oracle \
    --include cubepick-reach-layout-4

Pinned to robohub commit e472b1a1e041. The verifier and the reference solution are not published.

No trials yet.