All of these issues could be dealt with by running the algorithms on edge computing and tinyML hardware, but that is easier said than done. These …
http://dlvr.it/TKPxwW