DTBird’s ability to detect and deter large raptors, particularly golden eagles, and to reduce the risk of collisions with wind turbines (WTG) was evaluated in the USA by H.T. Harvey & Associates for the American Wind Wildlife Institute.

The evaluation was carried out in 7 DTBirdV4D4 units installed in 2016 in a California wind farm, using eagle-like unmanned aerial vehicles (aka, UAV) with a 2 m wingspan, and studying in situ large raptor behavior registered during 2017.

The evaluation yielded the following major results:

  • DTBirdV4 Detection Module:
    • Up to 99%1 Overall eagle-like UAV Detection rate from > 230 m to the Rotor Swept Area (RSA).
    • 63% mean Detection rate per distance band of 30 m (7 distance bands from >230 m to the RSA).
    • 2.5 False Positive rate (flight recording of something other than birds)/WTG/day2.
  • DTBirdD4 Collision Avoidance Module:
    • Deterrence response rates3 when considering unequivocally successful deterrence and potential deterrence:
      • 52-83% for golden eagles.
      • 36-76% for buteos.
      • 39-78% for all raptors. 
  • DTBirdV4D4 service reliability4:
    • >98% System.
    • >98% HD Cameras.
    • 100% Collision Avoidance.

To see the original report with DTBird notes, please go to this Link.


DTBirdV4D4 unit installed at Manzana Wind Power Project.

1 DTBird calculation for a flight path from >230 m to the RSA, through 7 distance bands with the mean detection rate presented in the Pilot Study (figure 19, page 51; DTBird note on the same page gives calculation details and online access to these calculations). Lower overall detection rates or even no detection is expected under specific conditions (e.g. flights entering the surveillance area at less than the maximum detection distance, or flight paths through sun glare areas).
2 The FP rate has been calculated by DTBird with the data provided in the Pilot Study, as detailed in DTBird note in epigraph: 4.1 DTBird Targeting Efficiency, page 58.
3 Including the potential interacting effect of the spinning blades, that requires a controlled experiment to be isolated.
4 Maintenance Summary provided by DTBird, located in table G1, pages 84-86.