Alright, science deals with particular sets of phenomena. Right? So, it deals with particular objects or things I see or perceive. For example, the phone, the bits and pieces of the phone, the weight, mass, charge, spin of the particles of the phone etc. etc.
Okay, there is no "universal" law that is holding these particular bits and pieces together. There are just the bits and pieces. The "whole" phone is just the sum total of the bits and pieces.
That's the problem of induction.
We don't have any experience of the universe in it's sum total, just the bits and pieces. We only assume we have a coherent picture of the universe in it's sum total, because of the regularity of those events. So every day we see that we can walk in a straight direction, so we assume that we can continue to do so. But there is no "thread" connecting the "pearls" of those bits and pieces, those events, all in-and-of-themselves. That's our inference.
Unless, you believe, like I do, that universals, natural laws, etc. exist. But the Modernist tradition is basically unanimously nominalistic. For reasons that are understandable, but nevertheless.
In other words, the problem is that we need to assume a "universal" constancy of a kind to be able to do science or live in the world at all, yet in pure empirical terms, no such universal can be known to exist. Because, by definition, empiricism only deals with particular objects and not universals.
Dunno if that made sense or not, but I tried.
Peace.