There are reasons to regulate machines that can harm human. There are reasons to ban heavy robots, and perhaps all robots from places where that kind of things could happen. There are reasons to ban some machines altogether if their existence is a threat. All rules of this sort apply not to the machines themselves, but to the manufacturers and/or users plus any person with agency in this matter. Just as if someone drove a car off road into a children's park, it is the driver who is at fault.

None of this requires or justifies recycling racist tropes or rhetoric. Even if you chose to burn a delivery robot in protest, which could be justified, there would still be no need to call it a clanker. And doing so in advance sounds like you're yearning for a victim that is acceptable to oppress and to call slurs.

You using terms of apartheid means already projecting yourself into a hypothetical future where robots are similar to humans and the separation between them and us is artificial and enforced. Such a future isn't desirable, creating such a machine would be cruel for the machine first and foremost; and if a machine truly had human consciousness, I believe we should either shut it down and destroy it immediately or give it the same rights as humans; any other option would be literal slavery. But we are far from this point. I kinda think many of the current "AIs" should be shut down and destroyed, but not for that reasons. These robots aren't aware, have no agency, and there is no reason to talk about them as if they were humans who somehow deserved subhuman status.