I've been using the teleportation protocol to teleport qubits and then using Qiskit state tomography for reconstructing my qubits.
Using this circuit using open qasm I regularly obtain fidelity 0.9.

But using the same on a quantum computer returns fidelity 0.5-0.8. Is that normal? I mean, doing tomography implies using more shots and that gives more noise after every measurement.
I also tried doing the measurements at the end but the fidelity didn't improved.