Agreed. This method of sustainable robots is generally not very good en mass. It's good for a hackaday post, but try making 100 of these. Before this is a possi

Agreed. This method of sustainable robots is generally not very good en mass. It's good for a hackaday post, but try making 100 of these. Before this is a possibility, we need robots that are able to self-model and hence understand the relationship between sensors and actuators correctly. Even 3D printing a robot on the same 3D printer, with the same calibrated parts from manufacturers, with the same software - _still_ requires calibration. The more complicated the robot, the more the errors accumulate. As for the quality of the paper - I completely blame the journal (who are ultimately responsible for the actions of the reviewers) as this should have been caught earlier. I can understand authors rushing to produce a paper not in their own language, but the profit machine journals need to take more responsibility for what they publish.