Connectionists: [PhD] Intelligent scarecrow

[Christos Dimitrakakis]

PhD Position in Avian Pest Control and Behavioural Modelling

We are looking for a PhD student interested in interdisciplinary research,
for modelling the behaviour of birds, and in particular corvids, in order
to design an adaptive AI system to minimise crop damage, i.e. an
intelligent scarecrow.
Essential information

• Host Institution: University of Neuchâtel and Agroscope
• Location: Neuchâtel and Posieux, Switzerland.
• Application Deadline: 30 November 2025. However, applications will be
reviewed as they come in.
• Start Date: January 2026 or later upon agreement.

To apply, send an email to christos.dimitrakakis at unine.ch with the subject
'PhD Intelligent Scarecrow' and the following documents:

   1.

   A CV.
   2.

   A one-page letter explaining why you are interested in (a) this project
   in particular (b) how you fit in with the rest of the resrarch group (c)
   your research interests more generally.
   3.

   Grade transcripts.

Project Background

Bird damage to crops remains a persistent issue, impacting yields and
economic viability, especially in maize and sunflower cultivation.
Traditional deterrents quickly lose efficacy due to habituation. This PhD
project is embedded in a national initiative to develop  intelligent,
autonomous bird deterrent systems that adaptively optimize stimulus
strategies. The project has three componnets.

1. Bird detection system: This consists of a set of prototype cameras
(hardware and software) capable of high-resolution imaging (64MP) with
real-time edge processing, which are used to detect bird activity.
2. Bird deterrence system: This consists of a small number of wireless
elements with acoustic, mechanical and optical actuators.
3. Adaptive deterrence algorithm: This is the core of the system and project,
which will aim to minimise bird presence over time.
 PhD Research Objectives

The PhD candidate will mainly contribute to the AI algorithm for analysing
bird behaviour and responding accordingly. This includes the following
tasks.

   -

    Development of avian behavioural and preference models, drawn from
   existing studies and our own field observations.
   -

    Development of centralised and distributed reinforcement learning
   algorithms, including model-free algorithms as a baseline, as well as
   model-based algorithms that take the avian models into account.
   -

   Development of a simulation environment for evaluating the reinforcement
   learning approach in silica.
   -

   Integration of the algorithm with the detection and deterrence platform,
   and testing, in collaboration with our partners in Agroscope.
   -

   Field deployment of the platform for in situ evaluation.

Candidate Profile

We are seeking a highly motivated researcher, willing to work in an
interdisciplinary team with:

   -

   Required: A Master’s degree in computer science, statistics, data
   science, artificial intelligence, robotics, electrical engineering, or a
   related field.
   -

   Required: Good background in AI/ML/statistics.
   -

   Experience in computer vision, embedded systems and/or field sensor
   networks is a plus, but not necessary.
   -

   Interest in applied ecological research and collaboration with industry.
   -

   Required: Fluency in English
   -

   Ability to work independently in lab and field environments.




-- 
Christos Dimitrakakis
https://sites.google.com/site/christosdimitrakakis
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/connectionists/attachments/20250904/89efaab5/attachment.html>