Martina Castellucci

Internship Neotron

MSc internship and thesis project: decision support for food contamination risk assessment & prediction.

Neotron logo
Neotron Laboratory (LMIB Division)
Part of the Cotecna Group

About Neotron

Neotron is part of the Cotecna Group, operating within testing, inspection and certification services with a strong laboratory network supporting multiple industry sectors.

Neotron
Testing & Analysis Food & Feed Pharma Data-driven labs

The “Where Data meets the Lab” approach highlights the integration between analytical workflows (high-throughput laboratory measurements) and data processing / bioinformatics to support decision-making.

Internship details

  • Host laboratory: Neotron Laboratory, LMIB Division
  • Start date: 02/02/2026
  • Scientific supervisor: Leila El Boulami
  • Track: MSc Bioinformatics — University of Bologna

Workspace

Office / lab environment

Lab + data workflows: from analytical measurements to structured datasets, dashboards and predictive models.

Thesis project — overview

Development of a Decision Support System for the Assessment and Prediction of Food Contamination Risk

A data-driven framework that combines regulatory limits, historical analytical data, and predictive modelling to support dynamic food safety monitoring.

Background

Food safety monitoring requires integrating analytical, environmental and regulatory data to detect contamination risks across food matrices. Contamination levels can vary seasonally and may be influenced by climatic conditions. The availability of large analytical datasets from Neotron enables computational approaches to improve risk prediction and decision-making.

Data

Neotron’s historical analytical database containing quantitative results across multiple food matrices and contaminant classes (concentrations, sampling dates, and metadata). External signals (e.g., climatic/seasonal indicators) may be integrated. Extraction and preprocessing are performed from structured SQL/CSV sources with confidentiality and data protection constraints.

Methods

  • Data cleaning, transformation and feature selection
  • Non-compliance estimation relative to EU legal limits
  • Temporal / seasonal statistical evaluation
  • Supervised ML models for contamination risk prediction
  • Dashboards / reporting with R and Power BI

Aims

  • Identify critical matrix–contaminant combinations
  • Detect seasonal/climatic patterns influencing contamination
  • Estimate future non-compliance probabilities
  • Deliver an integrated decision support framework for preventive action

Tools & deliverables

Python SQL / CSV Machine Learning R Power BI Reproducible workflows

Expected outputs include data harmonization pipelines, interpretable metrics on compliance trends, predictive models and a dashboard/reporting layer for decision support.