1. Soft Sensing of Oven-Cavity Moisture
Electrolux is seeking an MSc thesis student to support the development of a soft sensing estimation system for domestic ovens. The work will involve conducting experiments on certain food types, acquiring data using an oxygen sensor that measures moisture volume in the oven cavity, and analysing the resulting signals to create a data-driven virtual model of the process, possibly with physics-informed techniques included. The student must be available for a period of on-site laboratory work. Any programming environment may be used, but the final solution must be transferred to MATLAB/Simulink with equivalent performance.
30% theory, 50% software, 20% testing
Requirements: Python, MATLAB
Contacts: Imola Fodor, Giovanni Iacca
2. Continual Learning for Food Doneness Estimation
Electrolux is seeking an MSc thesis student to support the development of Continual Learning methods for an AI-based model, capable of estimating time-to-end of cooking processes. The work will investigate how the model can adapt over time to user preferences while maintaining reliable performance. The thesis can be carried out fully remotely. MATLAB is not mandatory during development, but the final algorithms must be delivered and validated in MATLAB/Simulink, achieving performance equivalent to the original implementation.
30% theory, 50% software, 20% testing
Requirements: Python, MATLAB
Contacts: Imola Fodor, Giovanni Iacca
1. Automated Model Selection and Hyperparameter Optimization for Time Series Forecasting
4MAINDS is looking for a thesis or internship student to develop a procedure for automatically selecting and configuring time series forecasting models. The work will focus on testing different approaches, from statistical models to deep learning and pre-trained foundation models, and finding suitable hyperparameter configurations. The results will be compared using appropriate validation strategies, with the aim of identifying the best model for each application scenario. The developed procedure will then be integrated into our energy monitoring and optimization platform.
30% theory, 50% software, 20% testing
Requirements: Python
Subjects: Machine Learning, Deep Learning
Contacts: Giacomo Segala, Giovanni Iacca
2. Investigation and Integration of Novel Time Series Forecasting Models
4MAINDS is looking for a thesis or internship student to explore new approaches to time series forecasting and integrate selected models into our platform. The work will start with a review of recent scientific literature to identify methods worth investigating. Selected models will then be implemented and tested against the forecasting solutions already available in our platform, using real-world energy datasets. Performance will be assessed not only in terms of prediction accuracy, but also considering computational requirements and inference time.
20% theory, 60% software, 20% testing
Requirements: Python
Subjects: Machine Learning, Deep Learning
Contacts: Giacomo Segala, Giovanni Iacca
There are various topics available at Oracle Labs. Topics of interest include (but are not limited to):
Automated Machine Learning with Explainability
Graph Machine Learning
Machine Learning for DBMS and Data Integration
Large Language Models for Semantic Search, Code Generation, and Assistants
Generative AI and AI Agents
10% theory, 70% software, 20% testing
Requirements: depending on the project
Contacts: Giovanni Iacca (will forward to the relevant team at Oracle)
There are various topics available at Biocentis, a company working on pest control. Topics of interest include:
Development of PesTwin GUI
Graph generator
Model calibration
Optimal release strategy
10% theory, 70% software, 20% testing
Requirements: see details here.
Subjects: machine learning, optimization, web development, data science
Contacts: Andrea De Antoni, Giovanni Iacca
There are various topics available at Aida Innovazione, a company working on AI and IoT applications. Topics of interest include:
Smart cities: optimization of traffic flows within car parks
Predictive maintenance for smart buildings (monitoring, control, and forecasting of plant energy consumption)
Fault/anomaly detection, preventive maintenance
Explainable AI for intelligent plant management
10% theory, 70% software, 20% testing
Requirements: Python
Subjects: machine learning, optimization
Contacts: Luca Cornali, Giovanni Iacca