To date we obtained 511 blue plates with multiple exposures of nine young clusters and associations to be used for studying the flare activity-age relation. First attempts showed already that automatic search increases the number of discovered flares by 50% compared to visual inspection (Aniol et al. 1989). Improved algorithms have now been implemented and applied. Automated object identification in existing catalogues (FSDB, GCVS, SIMBAD) is also available.

All plates are digitized with the PDS 2020 GMplus microdensitometers at AIM and reduced with the AIM software. Steps following the reconstruction of images include the transformation of plate to sky coordinates and matching of the multiple exposure chains with the stars on the deep reference plate. All complete chains (generally six exposures) are used to derive average light curves for all objects. For each magnitude interval 10-11, 11-12, … 15-16 the mean deviation and standard error of an image from the overall mean value of the respective chains is determined. If any image in a chain differs by more than a preset multiple of the mean error it is considered a potential flare event and registered for visual inspection, together with the chains immediately preceding and following the suspected flare event, to obtain information on the total duration of the outburst. The candidates are filed and the data of all stars are stored for checks of long-term variability.