Identifying diatoms is not without its problems. Apart from the obvious difficulties in the interpretation of identification keys (see above), at present the main problem areas are as follows:

1. Time spending:
Diatoms are used in a large number of scientific applications, mostly revolving around the reconstruction of environmental conditions, both past and present (e.g. palaeoclimate reconstruction). Associated with this is a large volume of routine diatom identifications carried out by the scientists involved, which can be extremely time-consuming and tedious. Individual samples can take hours to analyse, and large studies often deal with hundreds of samples at a time. Given the time pressure now commonly associated with short term scientific projects, time-savings on routine identification work would be widely welcomed.

2. Subjectivity of the identification process:
Identifying diatoms usually involves some interpretation of both images and species descriptions in identification keys. This can lead to personal "views" of diatom taxa which are not necessarily congruent with those of other workers, and may unbalance the outcome of some studies.

3. Training of new diatomists and non-specialists:
In some cases, the identification work is carried out by non-specialists for whom diatoms may only be a peripheral aspect of  their work, and who cannot reasonably be expected to have the same degree of expertise as a full-time diatomist. Also, newcomers to the field are often overwhelmed by the morphological diversity they are faced with.

4. Lack of identification literature:
There are now only few full-time diatom taxonomists world-wide, a trend which has not benefitted the field in terms of dealing with the staggering diversity of this fascinating taxonomic group. As a consequence, modern identification keys are far and few between, with the greatest lack occurring in the marine pennate diatoms (where most of the biodiversity appears to be locked up). Much of the existing literature was published in the last century or early this century and is consequently unavailable now. There are also language implications, with many of the more recent titles being unavailable in English.

There is thus an overwhelming case for new approaches to diatom identification. An automatic identification system such as that being developed in the ADIAC project will greatly alleviate all of the problems mentioned here. Time savings could be potentially enormous, identification can be made completely objective, and problems with availability and accessibility of printed literature should lose at least some significance. In addition, it could constitute a valuable training aid for those beginning to explore the world of diatoms.