In the general case, I guess. But watching gauges and dials like battery capacity only take a little work with a deterministic computer vision library.
People are making a concerted effort to force your business to do something, and you don't want to know their names or how much influence they actually have?
Gemma 4 E4B and Qwen 3 4B are pretty good, but fine-tuning makes them really good. There are tradeoffs at this size, so you'll have to find (or make) a finetune that does what you need.