While Aquila Ecology's AI cannot completely replace ecologist work, it can provide huge cost savings in the work required for Species Management Plans (SMPs). It also saves the ecologist the tedious work: figuring out hours of recording, or posting in places to conclude there are no bats. Below are questions and answers on how Aquila Ecology's AI can help with your SMP.

How can Aquila Ecology's AI analysis help with SMP research

  • Aquila Ecology can analyse huge datasets for a fraction of the cost of analysing by hand.
  • This can facilitate the data analysis that is necessary anyway in an SMP.
  • It also allows for the collection of additional data, for example by using standalone recorders to identify in advance the locations with high or, on the contrary, no bat activity, thus allowing for much more targeted research.

How does the analysis work

  • You can send or deliver the data on a hard disk
  • In general, we analyse the data within 3 working days
  • We can analyse both batexplorer, batlogger, and audiomoth data
  • Data provided with GPS location (batlogger data) can be delivered back as GIS-readable format
  • It is also possible to return the data in such a way that it is readable with BatExplorer
  • Finally, it is also possible to outsource data collection to Aquila Ecology. We have experience in placing and reading sensors suitable for this purpose.

How accurate is the analysis

  • The AI model distinguishes ambient noise (including pulses from electric cars .etc) from bat sounds with an accuracy of about 94%
  • The AI model has high accuracy for most common species: Common dwarf bat (98%), Rottish bat (96%), Lake bat (93%), Water bat (93%), Shaggy dwarf bat (77%), Common bats (73%), Common bats (73%)
  • It is possible to accept species only if they are predicted with >70% certainty. This greatly increases the accuracy
  • As a result, initial sorting with AI quickly saves 90% of valuable time

What are the costs

  • Cost of analysis is basically 250 for an analysis plus about 3 per GB
  • For very large data sets and projects, we discuss the price separately