Public results · robot arm

5 robots × 4 imSim standard tests.
20 verdicts published with video and evidence.

Select a test in the table to open its video and the reason for its verdict.

Robot \ test

Position accuracy

 

Pick and place

Fixed position

Pick and place

Find by itself

Part insertion

 

Panda

Franka

UR10

Universal Robots

PiPER

AgileX

Gen3

Kinova

TM12

Techman

Your robot · your AI

The same 4 tests, scored by the same rules

Pass
Fail
Caution · Inconclusive
Numbers: key value per test (mm, trials meeting the criteria, successes/measured)

All values are baselines. All robots were moved by the imSim baseline controller (not the manufacturer's controller or an AI), so this is not a performance ranking of robot manufacturers. Results in other domains are delivered to the requester in the same format. Test method

PiPER · Pick and place — find by itself at random positions — Pass 20/20

AgileX PiPER · 6 axes · reach 0.626 m · payload 1.5 kg

Pick and place — find by itself at random positions

Ask about this test

Trial 2 · Success

0:00 / 0:14

35 trials

Success 20
Excluded from measurement 15
Outlined square: video available

Pass

20/20

Successes / measured

Reason

20 of 20 measured trials succeeded. The lower bound of the 95% confidence interval of the success rate, 83.8%, is above the pass threshold of 80%, so the result is Pass.

Of 35 planned trials, 15 were excluded from the count because of ‘Position out of robot arm reach’. They count as neither success nor failure.

Success rate 95% confidence interval verdict
Rule version v2.3b-2026-10-04 · measured 2026-10-11

Verdict evidence

Measured values and criteria

Item

Measured value

Criterion

Results

Successes

20 / 20

Successes among measured trials

For reference

Success rate

100.0%

Point estimate (for reference)

For reference

Success rate 95% confidence interval

83.8% ~ 100.0%

Lower bound > 80.0% = Pass · upper bound < 40.0% = Fail

Pass

Excluded from measurement

15 / 35

Positions out of robot arm reach: 15

Excluded from the count (counted as neither failure nor success)

Minimum measured trials

20

4 or more

Met

Same test, 5 robots

95% confidence interval of the success rate (bar) and the success rate (dot). Dashed line: pass line 80% — Pass if the lower bound of the interval is above this line.

50%

60%

70%

80%

90%

100%

Test conditions and limits

Measured

Success rate of the robot finding an object at a random position within a defined range with its camera, grasping it and placing it at a defined spot — ground-truth object position used only for scoring (not for motion)

Placement

Random positions within a defined range

Trial

Planned 35 · measured 20 · excluded 15(Position out of robot arm reach 15)

Robot control

imSim baseline controller (neither the manufacturer's controller nor an AI)

Robot model

Checked against the manufacturer's published specifications (not against real robot behavior)

Test environment

In-house simulation test environment (not recognized by an accredited test body)

What this result does not show

  • Finding objects outside the defined range
  • Re-recognizing object pose in the hand
  • Grasp detection at the level of a real robot force sensor
  • Real robot performance — these are simulation results and do not replace physical tests
  • Performance ranking of robot manufacturers — this is a baseline produced with the imSim baseline controller
  • Official third-party approval — this is self-scoring in an in-house test environment

Also recorded

Not used for the verdict

Pass and fail are decided only by the evidence above. The items below are recorded for every trial and published as measured.

No failures or cautions among the evaluated items.

The same test with your robot · your AI

Same verdict rules, same result format — video, verdict, evidence.

Get started