is pretty clear implication. But I'll spell it out for you - people with impairment to senses (primarily vision) rely on this infrastructure to know when it's safe to cross a road.
These devices will also often indicate their location when you hold the button. That can help a visually impared person confirm that they're oriented correctly. That use case is likely why these are field programmable; factory programming would be sufficient for limited prompts (wait, begin crossing, count down, clicks and whistles, etc) but not for street names.
