If your timekeeper doesn't perform adequately, there's an impressively overengineered neural net project to read clocks: https://arxiv.org/abs/2111.09162 "It's About Time: Analog Clock Reading in the Wild".
(Although I don't know how it would compare to the latest generalist vision models... Clocks are pretty well labeled in photographs online, one would expect.)
(Although I don't know how it would compare to the latest generalist vision models... Clocks are pretty well labeled in photographs online, one would expect.)