Yeah that's a great question!

How do we generate the pairs? We use an entanglement protocol called Single-Click-Entanglement (you can read details about it here: https://www.nature.com/articles/s41586-018-0200-5), essentially we make the node emit a photon that's entangled with our communication qubit. A second node does the same. We bring the two outputs on a beam-splitter, and if we detect a photon on the other side we have entangled the two communication qubits. Most of the time the photons get lost along the way, and the attempt fails, but once every while we get a click from the photon detector (hence the name of the protocol) that tells us that an entangled state has been made and we can stop trying.

How do we verify entanglement? Once the protocol succeeds, we know we have an entangled state. To characterize it we use a measurement procedure called quantum state tomography, that by measuring the same state many times in different ways can tell how good it is.

How long do they last? In our experiments they last a few tens of a second, which is very long compared to other quantum platforms, and compared to how long it takes to do operations on the qubits (hundreds of nanoseconds).

What can you do with it? Entangled states are at the basis of many (all?) quantum network applications. Just like for quantum computers, we don't really know yet all the things we'll be able to use it for. You can find many more details here: https://www.science.org/doi/10.1126/science.aam9288