I’ve experienced this during one of my internships. One half of the team was set to work doing programming for some output data from a technology no one had experience with (RFID chips), and the other half’s goal was to get that technology working. Unfortunately about halfway through the project timeline it became apparent that the RFID chips would not be working out to the extent needed so our project manager needed to do a complete turnaround. However, considering that testing the feasibility of this technology was one of the side goals of the project in the first case, it wasn’t a complete failure.
I think these specific issues could have been avoided with more research processes done beforehand, but it seems to me these kinds of issues pop up regardless of how much research is done.