XMM-Newton has produced an extensive X-ray source catalogue in which the standard pipeline determines the variability of sufficiently bright sources through chi-square and fractional variability tests. Faint sources, however, are not automatically checked for variability, thus overlooking faint, short timescale transients. Our goal is to find new faint, fast transients in XMM-Newton EPIC-pn observations. To that end we have created the EPIC-pn XMM-Newton Outburst Detector (EXOD) algorithm, which we run on the EPIC-pn data available in the 3XMM-DR8 catalogue. In EXOD, we compute the whole-field variability by binning in time the counts in each detector pixel. We next compute the maximum-to-median count difference in each pixel to detect variability. We applied EXOD to 5,751 observations and compared the variability of the detected sources to the standard chi-square and Kolmogorov-Smirnov (KS) variability tests. The algorithm is able to detect periodic and aperiodic variability, short and long flares. Of the sources detected by EXOD, 60-95% are also shown to be variable by the chi-square and KS tests. We obtain a net number of 2,536 variable sources. Of these we investigate the nature of 35 sources with no previously confirmed classification. Amongst the new sources, we find stellar flares and AGNs; but also four extragalactic type I X-ray bursters that double the known neutron-star population in M31. This algorithm is a powerful tool to promptly detect variable sources in XMM-Newton observations. EXOD detects fast transients that other variability tests classify as non-variable due to their short duration and low number of counts. Finally, EXOD allows us to detect and identify the nature of rare compact objects through their variability. We demonstrate this through the discovery of four extragalactic neutron-star low mass X-ray binaries, doubling the number of known neutron stars in M31.