Artificial potential barriers are known that allow robots to avoid obstacles. They are constructed as follows. https://authors.library.caltech.edu/106548/1/2010.09819.pdf
Can you please tell me how this approach is generalized to the case of limiting several coordinates of the controlled robot?