.png)
Location: Delft, NL
Duration: 4–6 months (flexible)
Start date: In consultation
About the assignment
We have developed a prediction model that forecasts caterpillar population spikes based on moth flight activity. By improving the timing of biological control, this model helps growers reduce their dependence on chemical crop protection.
However, the current model does not yet incorporate environmental factors like temperature and humidity, which play a critical role in how quickly moths develop into caterpillars. Biologically, the development rate from moth → egg → caterpillar is strongly influenced by temperature (and to a lesser extent humidity). Recently, we have integrated temperature and humidity sensors into our systems. This opens up the opportunity to significantly improve prediction accuracy.
Objective
The goal of this internship is to enhance our existing prediction models by incorporating temperature and humidity data, leading to more accurate and timely forecasts of caterpillar outbreaks.
What you will do
You will work on extending and improving our predictive models, including:
● Analyzing existing datasets (moth flight events + environmental data)
● Investigating how temperature and humidity influence development cycles
● Designing and implementing improved prediction models (e.g. degree-day models, time-series models, or ML approaches)
● Comparing model performance against the current baseline
● Validating results using real-world data
● Iterating and optimizing model accuracy and robustness
● Collaborating with the team to integrate the improved model into production systems
What we are looking for
● You are enrolled in an HBO or WO program (e.g. Data Science, AI, Applied Mathematics, Biology, Biosystems Engineering, or similar)
● You have experience or strong interest in:
○ Data analysis and modeling
○ Python (or similar)
○ Time-series or predictive modeling
● You are comfortable working with real-world, imperfect datasets
● You think analytically and are able to translate theory into practical solutions
● Affinity with biology, horticulture, agriculture, or ecology is a plus
What we offer
● A challenging internship at the intersection of biology, data science, and product development
● Direct impact on real-world agricultural outcomes
● Access to unique datasets combining sensor data and biological events
● Freedom to experiment with different modeling approaches
● Guidance from an experienced technical team
● Internship compensation
Interested?
Send us your CV and a short motivation. We’d love to hear how you would approach improving these predictions.
