Under the title 'RoboCup 2021 Worldwide', the traditional competition was held in an unusual form due to the pandemic: from June 22 to 28, 2021, the teams came together exclusively virtually to compete against each other and face various challenges. B-Human - the team from the University of Bremen and the German Research Center for Artificial Intelligence (DFKI) - won three of the four new competitions in the Standard Platform League and successfully defended its 2019 title.

For the first time, the RoboCup World Cup was not about the number of goals scored. Instead of a traditional soccer tournament, the international teams were faced with competition formats that could not fully replicate the 'real' competitions, but in some cases presented new technical and scientific challenges. In the Standard Platform League (SPL) of RoboCup 2021, the regular competition, in which teams of five NAO robots normally compete against each other, was replaced by four individual competitions.

The first two competitions - the 'Passing Challenge' and the 'Obstacle Avoidance Challenge' - focused on the robots' soccer skills and were each played with their own NAOs on their own pitch. In the 'Passing Challenge', two robots had the task of passing a ball past obstacles as often as possible within five minutes. With a total of 27 passes between the Bremen NAOs, B-Human achieved by far the best result in this competition. The aim of the 'Obstacle Avoidance Challenge' was to dribble the ball past several obstacles into the goal as quickly as possible. Here too, B-Human won the competition as the fastest team.

The far more interesting and challenging competitions from a scientific point of view awaited the teams with the 'Autonomous Calibration Challenge' and the '1 vs. 1 Challenge'. Here, they transferred their software via the Internet to other teams' laboratories to run it on other NAOs and prove themselves on other teams' soccer pitches. The 'Autonomous Calibration Challenge' was about finding the positions of two balls on the field in the shortest possible time and approaching three given positions as precisely as possible. In a preceding 'calibration phase', the software automatically makes adjustments to the previously unknown robot. The Bremen software tackled this complex task in three different laboratories - in Dortmund, Amsterdam and Hamburg - and achieved the best score in each case. Other teams were sometimes slightly faster, but B-Human managed to determine the most precise positions in almost all cases.

The '1 vs. 1 Challenge' was the only competition that was actually played in matches against each other. As this was also played on foreign pitches and foreign robots, the knowledge and skills gained from the 'Autonomous Calibration Challenge' were of great value. B-human competed against the NAO team HTWK from the Leipzig University of Applied Sciences. Both teams were similarly fast and determined, but B-human missed the opponent's goal several times, meaning that they were unable to make up the resulting narrow deficit.