Context. The quantitative study of the physical properties and chemical abundances of large samples of massive blue stars at different metallicities is a powerful tool to understand the nature and evolution of these objects. Their analysis beyond the Milky Way is challenging, nonetheless it is doable and the best way to investigate their behavior in different environments. Fulfilling this task in an objective way requires the implementation of automatic analysis techniques that can perform the analyses systematically, minimizing at the same time any possible bias. Aims: As part of the ARAUCARIA project we carry out the first quantitative spectroscopic analysis of a sample of 12 B-type supergiants in the galaxy NGC 55 at 1.94 Mpc away. By applying the methodology developed in this work, we derive their stellar parameters, chemical abundances and provide a characterization of the present-day metallicity of their host galaxy. Methods: Based on the characteristics of the stellar atmosphere/line formation code fastwind, we designed and created a grid of models for the analysis of massive blue supergiant stars. Along with this new grid, we implemented a spectral analysis algorithm. Both tools were specially developed to perform fully consistent quantitative spectroscopic analyses of low spectral resolution of B-type supergiants in a fast and objective way. Results: We present the main characteristics of our fastwind model grid and perform a number of tests to investigate the reliability of our methodology. The automatic tool is applied afterward to a sample of 12 B-type supergiant stars in NGC 55, deriving the stellar parameters, Si , C , N , O and Mg abundances. The results indicate that our stars are part of a young population evolving towards a red supergiant phase. For half of the sample we find a remarkable agreement between spectroscopic and evolutionary masses, whilst for the rest larger discrepancies are present, but still within the uncertainties. The derived chemical composition hints to an average metallicity similar to the one of the Large Magellanic Cloud, with no indication of a spatial trend across the galaxy. Conclusions: The consistency between the observed spectra and our stellar models supports the reliability of our methodology. This objective and fast approach allows us to deal with large samples in an accurate and more statistical way. These are two key issues to achieve an unbiased characterization of the stars and their host galaxies. Based on observations obtained at the ESO VLT Large Programme 171.D-0004.