EPR spectroscopy is intrinsically complicated. This is not only due to the underlying physics, but due to mostly overlapping broad lines. Individual contributions can usually not simply be disentangled, and often, the only chance to extract parameters from spectra is by fitting spectral simulations. Compare this with the relative ease of interpreting NMR or optical spectra. Additionally, setting up experiments and choosing the appropriate parameters is usually left to the operator and requires experience and thorough understanding of both, hardware and underlying physics.
There are two levels of complication in EPR spectroscopy that need to be distinguished: choosing appropriate experimental parameters and analysing the resulting data. However, both can be automated to a great extent, and lack of automation and routine procedures is responsible for the apparent complexity of EPR spectroscopy as well as its slow pace and (partly) its unreliable results.
Both aspects have been solved for some rather special routine applications, and there are a few fields where EPR spectroscopy is used in routine analysis, even in industry. However, cutting-edge research tasks often require a complete different approach and much more control over both, setting experimental parameters and data analysis and parameter extraction.
Many experimental tasks can and should be automated, and even if they are not implemented as standard protocols in software, they can be applied and carefully followed by experimenters. Therefore, a first step is to thoroughly train the people using the spectrometers, starting with students. A second step is to implement and automate all experimental steps possible. This will often require using own software, as will be detailed in the tools section.
The same applies to data processing and analysis. Already reproducibility of data processing and analysis – i.e. meeting fundamental standards of science – requires using software and tools that help the user log the individual steps. Additionally, most of the tasks necessary to extract parameters from EPR spectra have long been solved, although the EPR community is not necessarily aware of it. What we are lacking is rigorously and consistently applying the best state-of-the-art tools from both, digital signal processing and optimisation. Integrating these tools in a modular fashion in a consistent and easy-to-use framework for spectral data processing and analysis is clearly part of the solution.