| System under test | Success Criteria | Result |
| LIDAR localization | Translational error <= 10 centimeters in the x and y axis | A translational component within the error region of 10 centimeters |
| Yaw error <= 10 degrees | Rotation component within the error region of 10 degrees | |
| Successful initialization with initial guess error deviation of >=20 centimeters and >=20 degrees with the ground truth | Successful initialization with initial guess deviation of 40 centimeters and 22.5 degrees | |
| Aruco3 marker detector | Translational error in range estimation under 3cm | Observed error value under 1cm |
| Rotational error in marker yaw estimation under 3 degrees. | Observed error value under 2 degrees | |
| Global planner | Should plan paths for multiple robots in under 2 seconds. | Decentralized planners. Plans new paths in 22ms. |
| Should work in simulated testbed env. | Works in testbed simulation. | |
| Collision Avoidance | Should avoid collisions 100% of the time | Avoids collisions attributing to replanning and inflation maps |
| 3D Object Detector | Should achieve an average recall and precision of 70% for objects within 5m | Average Recall (AR) = 89.1 % Average Precision (AP) = 71.1 % |
| Docking | Should dock within a lateral error of 10cm, a forward error of 2cm and orientation error of 10 degrees | Lateral error <= 5cm Forward error <= 2cm Orientation error <= 10 degrees |
