EuRoC Forum

[lrubio]

I think these modifications could have been discussed a little before than 2 days before to final submit, because some of us don't have too much time to change more code. We understood that mobile base was needed in task3 in order to be able to pick and place far objects, but not 'illegal' to use in task1-2 because real robot always have a mobile base.

I would agree with that decision, because is other way to understand these task descriptions, but not 2 days before to final submit when we are in extended deadline.

Regards,
Luis Rubio

[lrubio]

Any answer would be grateful

[borst]

Dear Luis, we appologize for the inconvenience caused by this bug in the simulator. We have designed Task 1-3 in a way that they test different abilities in programming robots. Task 1 is just pick and place with quite precise object locations, Task 2 adds more noise to the object localization and the robot precision while Task 3 was aimed to use the base position of the robot in planning as well. That way our "legal" document (the specification of the chalenge) was written and approved by the EC. So changes affecting this document cannot be done easily. Although legally almost impossible we have discussed the consequences if we would accept base movements in Task 1 and 2. I think you agree that using the base for example to reach always the same relative position for a grasp approach simplifies task 1 and 2 significantly. That means the teams investing time in special solutions to cope with different grasping situations (which will occur in the real setup as the base cannot be moved in any configuration cause of its own geometric extensions) would have a disadvantage. So in any case some challengers have drawbacks.
We applologize but we were not aware of the bug and therefore reacted on the announcement immediately.
At the end we hope that still all teams qualify for the match making meeting. After that the socores and the proposal for an end user scenario will both be counted to get to the stage 2. We hope this bug issue will play no role in the decision for the winning 5 teams.

[lrubio]

Dear Luis, we appologize for the inconvenience caused by this bug in the simulator. We have designed Task 1-3 in a way that they test different abilities in programming robots. Task 1 is just pick and place with quite precise object locations, Task 2 adds more noise to the object localization and the robot precision while Task 3 was aimed to use the base position of the robot in planning as well. That way our "legal" document (the specification of the chalenge) was written and approved by the EC. So changes affecting this document cannot be done easily. Although legally almost impossible we have discussed the consequences if we would accept base movements in Task 1 and 2. I think you agree that using the base for example to reach always the same relative position for a grasp approach simplifies task 1 and 2 significantly. That means the teams investing time in special solutions to cope with different grasping situations (which will occur in the real setup as the base cannot be moved in any configuration cause of its own geometric extensions) would have a disadvantage. So in any case some challengers have drawbacks.
We applologize but we were not aware of the bug and therefore reacted on the announcement immediately.
At the end we hope that still all teams qualify for the match making meeting. After that the socores and the proposal for an end user scenario will both be counted to get to the stage 2. We hope this bug issue will play no role in the decision for the winning 5 teams.

Thanks for the answer,
Luis