Just got back from running a two day course on sound analysis in the New Forest. It was a good opportunity to review the literature on automated logging of bat passes and automatic recognition of bat calls. There are some very interesting statistical techniques being used, but I have to say, not with a great deal of success, mostly with separating some Myotis species and especially with the separation of serotine from Leisler's. There is also considerable confusion between Plecotus and Barbastellus. Close but no cigar.