Dear SSL Teams,

There have been some complaints about the acceleration and max_speed in the simulation. While we intended to model the fact that we all have real robots and all robots are different, we noticed that the accelerations and speeds are not what we expected. Some teams that performed very well in terms of acceleration and speed during the last physical RoboCup in Sydney seem to be ending up at the slow side in this year's simulation.

This could be due to improper specification regarding the kind of acceleration we meant (perfect conditions or something you can reach reliably), or teams could have copied others if they had no idea how to get these values or weren't able to go to their lab to determine them. Furthermore, we as the TC/OC might be to blame for not coming up with a good scalable way of checking the values you submitted, and teams preferring overestimating themselves over underestimating.

Whatever it might be, the TC/OC decided that we will switch to a common acceleration profile, starting from the lucky losers and knockout matches. The common accelerations will be the average of all submitted specs, i.e:

acc_speedup_absolute_max: 4.7,

acc_speedup_angular_max: 50.8,

acc_break_absolute_max: 5.9,

acc_break_angular_max: 51.6,

vel_absolute_max: 3.7,

vel_angular_max: 26.2

We apologize for the inconvenience.

Carla Cosenza

RoboCup Small Size League | Organizing Committee