en entrée

19:56 Sep 27, 2017 

Many thanks to both - looking at the context I think it may refer to input of data on anomalies into a file

19:47 Sep 27, 2017 

I think Nathalies "at input" is possibly slightly more applicable than your own suggestion of "incoming" — after all, it is not unreasonable to imagine in any system that many 'anomalies' would have been filtered out on an incoming 'stream'; however, despite the fact they are not in themselves anomalous, certain situations might prove to be an anomaly at the moment of input: for example, the correct part, free from faults, arrives at the input too early / too late.

19:13 Sep 27, 2017 

Thanks Nathalie - it might be! It's one of those vague unspecific docs where you have to try to work out what they are talking about.