Nemo-simulator

Context

NEMO project (Marie Curie Fellowship) in collaboration with INRAE and Institut Agro. The project aimed at delivering operational decision-support tools for managing nematode pests in potato cropping systems.

Problem

Plant-parasitic nematodes significantly reduce potato yields. Farmers and advisors need tools that translate soil counts, virulence levels and crop rotation options into clear recommendations (variety choice, rotation schedule, expected yield loss).

Solution

Nemo-simulator is an interactive web simulator combining:

• mechanistic epidemiological models of nematode population dynamics;
• parameterization from field data (soil counts, soil types, varieties);
• scenario testing for rotations and interventions (cultural control, biocontrol);
• a user-friendly interface to advise and visualize field-scale outcomes.

The tool provides temporal charts, risk maps (if geolocation data is available) and a report export module.

Technologies

• Language: Python
• UI Framework: Streamlit
• Models: mechanistic modelling (ODEs/discrete models), parameter calibration using optimization (e.g., differential evolution, MCMC when needed)
• Visualization: Plotly / Altair inside Streamlit
• Demo hosting: Streamlit Cloud (or dedicated server)

Impact

• Reduced chemical treatments through improved rotation planning.
• Yield improvements observed in pilot fields.
• Tool used operationally by advisors for decision making.

Links & Access

• Public demo: https://nemo-simulator.streamlit.app/
• Contact: israeltankam@gmail.com