An ancient rolled up and buried Roman tablet was found in Kent, England. A group of researchers attempted to read the tablet without unrolling it by using neutron computed tomography, however this method failed as the resolution was not high enough to determine what was written on the tablet. For this reason, the researchers decided to carefully unroll the tablet and use a scanning electron microscope to read it. This approach was successful and finally allowed the researchers to decipher the text. The article describing this can be found here.
This is a great example of how different technology is suitable for different tasks in archaeology. Although the neuron computed topography method may have worked for other archaeological problems, in this case it was not suitable. The researchers tested it, it failed and then they moved on to a new method.
When the tool or computational approach that you are using is not doing what you need it to do, consider using a different one as, often, you may find that what you are using is not suitable to the problem you are solving. In the above article the first approach was not suitable, but then the researchers decided to use a scanning electron microscope instead the results were positive.