P22/23-ENG

Category: P22/23-ENG

Carbon farming: future perspectives and the contribution provided by…

July 22, 2026
by admin

Climate change is one of the most pressing challenges of our time, having significant impacts on ecosystems, agriculture and food security. Carbon farming is gradually emerging as one of the solutions to counter climate change in agriculture through the implementation of practices that reduce emissions and increase carbon sequestration in the soil. This approach can also generate an innovative business model, financially incentivising farmers to adopt such practices, thus contributing to a more environmentally sustainable agriculture, with possible further social and economical benefits. This thesis aims to analyse the state of the art of carbon farming projects in the world and explore how digital innovation plays a role within carbon farming projects and how they support carbon farming business models.

Required competencies: keen interest in the agri-food sector and digital innovation in the sector; ability to meet deadlines set with the tutor; proactivity and perseverance in achieving set objectives

Docente di riferimento Prof. Cosimo Pacciani (cosimo.pacciani@polimi.it)

P22/23-ENG

Computer Vision for Weed Detection

July 17, 2026July 17, 2026
by admin

This thesis focuses on the development of a vision-based system for automatic weed detection using an onboard RGB camera. The goal is to enable reliable discrimination between crops and weeds in real time through deep learning-based perception, providing actionable feedback to the robot navigation and control system for targeted intervention. The project investigates single-shot and few-shot learning strategies to reduce the need for large annotated datasets, improving adaptability across different field conditions and crop scenarios. The expected outcome is a robust perception module that supports precision agriculture tasks such as selective weeding and intelligent field coverage. The project requires Python3 and deep learning frameworks (PyTorch). Dataset selection and evaluation metrics will be part of the work.

Relevant references:
– https://arxiv.org/pdf/2304.05832
– https://ieeexplore.ieee.org/document/10182963

Reference teacher: Prof. Gianni Ferretti – gianni.ferretti@polimi.it

P22/23-ENG

Terrain Segmentation and Crop Classification from Aerial Images

July 17, 2026July 17, 2026
by admin

This thesis focuses on large-scale agricultural analysis using aerial or satellite imagery.

The main objective is to develop a system capable of classifying crop types from remote sensing data. The project can be extended in several directions:

Leveraging temporal data to predict optimal crop planning for future seasons
Integrating farm boundaries to generate crop rotation strategies
Estimating cultivated area per crop type across multiple farms

The implementation will rely on Python3 and machine learning/deep learning methods for semantic segmentation and classification.

Relevant references:

Reference teacher: Prof. Gianni Ferretti – gianni.ferretti@polimi.it

P22/23-ENG

Fruit Detection and Maturity Estimation

July 17, 2026July 17, 2026
by admin

This thesis addresses computer vision techniques for fruit detection and ripeness assessment.

The student will develop a system that:

Detects fruits within images using deep learning models
Classifies their maturity stage based on visual features

A possible extension includes implementing a regression model to estimate the remaining time until harvest for each detected fruit.

The project requires Python3 and deep learning frameworks (PyTorch). Dataset selection and evaluation metrics will be part of the work.

Relevant references:

Reference teacher: Prof. Gianni Ferretti – gianni.ferretti@polimi.it

P22/23-ENG

Tractor Path Coverage Optimization

July 17, 2026July 17, 2026
by admin

This thesis focuses on coverage path planning for agricultural vehicles.

The objective is to develop a software system that, starting from recorded robot trajectories, generates a field representation suitable for coverage planning. The system should integrate with tools such as Fields2Cover or the Nav2 coverage server to compute efficient coverage paths.

Key tasks include:

Processing recorded paths into geometric field representations
Generating coverage strategies using existing planners

Implementation can be done in C++ or Python3. The project can be conducted entirely in simulation, with optional real-world validation at the Cremona’s campus or nearby test fields.

Relevant resources:

Reference teacher: Prof. Gianni Ferretti – gianni.ferretti@polimi.it

P22/23-ENG

Advance Estimation of Biochemical Methane Potential (BMP) Using Machine…

June 3, 2024July 16, 2026
by admin

Biochemical Methane Potential (BMP) testing is an essential tool for deriving practical knowledge for optimizing and operating large-scale anaerobic digesters, monitoring, modeling and evaluating process performance, or when it is ongoing the development of a scenario analysis. While its usefulness is undoubted, the long duration of BMP testing is problematic for many of its applications, especially when timely results are required for decision making. In recent decades, numerous scientific contributions have demonstrated that a reduction in BMP test duration is possible by predicting the final gas production. The aim of this thesis work is the development of a new procedure/algorithm to obtain a preliminary estimate of the BMP experimental result using Machine Learning. The effectiveness and efficiency of the developed algorithm will be verified using experimental data from BMP tests performed on different substrates commonly fed to anaerobic digesters.

Reference teacher: Prof. Gianni Ferretti – gianni.ferretti@polimi.it

P22/23-ENG

A robot for the automation of the stable

April 30, 2024April 30, 2024
by admin

Barn automation is gradually becoming more and more widespread. Among the solutions offered by the market we find, for example, mobile robots for cleaning stables and distributing feed. However, these solutions are characterized by a low level of automation and, more generally, by extremely simple technological solutions, which make them robust but not very flexible.

This work concerns the design and implementation of an autonomous mobile robot, based on the commercial Agilex Bunker/Bunker Pro robot, for feed distribution. Compared to current commercial solutions, the robot must be able to navigate autonomously both inside and outside the stable, without requiring dedicated infrastructure (easy routes, special localization devices, etc.).

This work is suitable for both students coming from an engineering and non-engineering background, and can be tackled both by a single student and by a group of two students.

Reference teacher: Prof. Luca Bascetta – luca.bascetta@polimi.it

P22/23-ENG

Estimate the slip of a track with machine/deep learning…

April 30, 2024April 30, 2024
by admin

Model-based control techniques are widely used for autonomous navigation, but achieving good performance in the presence of disturbances and uncertainties requires online estimation and adaptation of the model. Agricultural applications are a typical example where the variability of the environment and the complexity of the wheel/terrain interaction model require an approach based on a simple model of the robot/vehicle, equipped with algorithms capable of continuously adapting/improving this model using all sizes available.

This thesis aims to develop classical and machine/deep learning techniques to estimate online the slip of a track, constituting the fundamental element for the creation of a controller for the autonomous navigation of a skid-steering robot for offroad/agricultural applications.
Different sensors will be taken into consideration, starting from the state of the robot and the information provided by an IMU, up to the images of the terrain generated by an RGBD camera or a lidar.
A key aspect of the thesis will be the comparison between classical and learning-based techniques, in order to discover any added value of the latter.
The developed algorithms will be validated in simulation and/or in the field using a tracked robot.

This work is suitable for both students coming from an engineering and non-engineering background, and can be tackled both by a single student and by a group of two students. Knowledge of Python is required.

Reference teacher: Prof. Luca Bascetta – luca.bascetta@polimi.it

P22/23-ENG

Localization and mapping for autonomous navigation in the vineyard

April 30, 2024April 30, 2024
by admin

Are localization and mapping necessary to autonomously navigate a vineyard?
Do we really need to create a map of the vineyard with sensors or can we rely on the known structure of the rows?
Do we really need a complex localization algorithm or can we just rely on GPS measurements?
Are the localization and mapping approaches that are usually adopted in robotics, autonomous vehicles or in the open field suitable for the vineyard, or can we use simpler approaches that are able to exploit, as much as possible, the structure of the rows?
Does using artificial intelligence help?

This work aims to answer one or more of the previous questions to develop simple, efficient and reliable localization and mapping techniques to autonomously navigate a vineyard.

This work is suitable for students coming from an engineering background and with experience in using the C, C++ or Python language. A group made up of two students, one from an engineering background and one from a non-engineering background, represents the best solution.

Reference teacher: Prof. Luca Bascetta – luca.bascetta@polimi.it

P22/23-ENG

Renewable energy production and efficient use at farm-level

January 11, 2023April 30, 2024
by admin

The renewable energy potential in the agricultural sector, at farm-level, is not yet fully exploited and opportunities are foreseen to develop interesting solutions in this area.
The thesis will analyze the feasibility of developing renewable energy technologies in the context of farms. Best practices and the most advanced renewable-based farm configurations will be reviewed at the EU level. Different case studies (in terms of farm scale, type and location) will be investigated at the Italian level (possibility to collaborate with a real farm to be verified). Energy storage solutions for reducing the effect of non-dispatchability of renewable sources, while increasing the self-use of electricity and heat at the farm scale, will be considered. Moreover, the efficient use of energy resources and the application of new energy technologies and digital tools will be studied.

Reference teacher: Prof. Manuele Gatti – manuele.gatti@polimi.it

P22/23-ENG

Recovery of nutrients from agricultural biowaste for sustainable fertilizers

December 12, 2022December 12, 2022
by admin

Population growth, changing in diet and lifestyle, modern agriculture and intensive livestock farming are the main drivers responsible for the increase in food production observed in the last 50 years, which reached approximately 24 million tons of food per day worldwide. The associated excessive use of chemical fertilizers in the agricultural sector causes high environmental impacts, such as eutrophication and greenhouse gas emissions, along with the exploitation of soil, water and natural resources. Furthermore, as a results of the agricultural activities, huge amount of waste is generated, with livestock manure as main contributor.
Thus, innovative sustainable solutions are required to reduce the pressure on finite resources, to decrease the amount of chemical input in agriculture, and to promote a more responsible management of waste. In the view of circular economy, re-use and valorisation of agricultural waste are of paramount importance to foster the production of bio-based fertilisers. In addition, waste treatment technologies, such as anaerobic digestion and hydrothermal carbonization, represent a solution to stimulate material and energy recovery.
Nutrients, such as phosphorus, nitrogen and potassium, can be recovered from livestock manure to produce high-quality fertilizer ingredients, as struvite and K-struvite. The purpose of the M.Sc. thesis is to investigate different strategies for nutrient recovery from such secondary source. Experimental activities will be performed at the lab-scale in the view of developing innovative technologies to scale-up in relevant environment in cooperation with industrial stakeholders. A possible recovery route will concern the recovery of nitrogen and phosphorus from the liquid fraction of digestate via selective precipitation, ion exchange or other separation/purification techniques. The recovery will be optimized to maximize the nutrients recovery efficiency and recovered products characteristics, while minimizing the addition of chemicals and the use of energy, to guarantee solution techno-economic feasibility and environmental sustainability.

Reference teacher: Dr. Andrea Turolla – andrea.turolla@polimi.it

P22/23-ENG

Resource recovery from agri-food biowaste by purple phototrophic bacteria

December 12, 2022December 12, 2022
by admin

Among different waste streams, agri-food biowaste has become increasingly more relevant due to the significant environmental consequences resulting from improper treatment and disposal. Resource recovery from such biowaste streams, turning waste into valuable resources (such as bioplastics, biofuels, single cell proteins and biostimulants), can represent a promising solution to reduce environmental impacts related to waste disposal and natural resource consumption.
The innovative application of purple phototrophic bacteria (PPB) for resource recovery from various biowaste streams, particularly as microbial mixed culture grown via photoheterotrophic metabolism, is drawing considerable attention. In fact, by using solar light as energy source, PPB can simultaneously assimilate carbon and nutrients at high efficiencies, while maximizing the recovery of resources.
As the use of PPB for high-rate biowaste stream treatment and resource recovery is still in its early stages, the primary goals of this project are (i) to improve the understanding of mechanisms of PPB growth on agri-food biowaste through a combined experimental and modelling approach, and (ii) to investigate the integration of resource recovery by PPB in existing plant layouts, in the view of enhancing process performance and promoting sustainable scale-up. To this end, lab-scale reactors for PPB cultivation will be designed and operated to identify the optimal operating conditions. Moreover, resource recovery will be evaluated, possibly looking at biopolymer accumulation in biomass and protein production. Finally, based on the experimental and modelling results, a pilot-scale optimized photobioreactor will be designed and developed.

Reference teacher: Dr. Andrea Turolla – andrea.turolla@polimi.it

P22/23-ENG

Localization and mapping for agricultural applications

December 2, 2022December 5, 2022
by admin

Localization and mapping are two fundamental parts of the navigation system of an autonomous robot. Mapping is the procedure that allows the robot, using the available sensors, to generate a map of the environment. Localization, instead, is the procedure that allows the robot, using its sensors and a map of the environment, to determine its position and orientation with respect to a fixed reference frame.
These two tasks, that are always not easy, are particularly complex when an outdoor agricultural scenario is considered, due to the light conditions, the seasonal variability, and the overall complexity of the natural environment.

This thesis aims at setting up and comparing the performance of different localization and mapping algorithms, or of simultaneous localization and mapping algorithms on some realistic agricultural datasets.

This work is more suitable for students with an engineering background and with a previous experience in coding using C, C++ or Python. A group of two students, one with an engineering background and one with a non engineering background would be the best choice.

Reference teacher: Prof. Luca Bascetta – luca.bascetta@polimi.it

P22/23-ENG

LFT modelling and identification of BMP tests

October 10, 2022December 5, 2022
by admin

The Biochemical Methane Potential (BMP) test is an essential tool to derive practical knowledge for optimizing and operating full-scale anaerobic digesters, monitoring, modelling, and evaluating process performances, or when a scenario analysis is being developed. Despite its usefulness is undoubted, long duration of BMP tests is problematic for many of its applications, especially when prompt results to take decisions are required. Over the last decades, several scientific contributions demonstrated that a reduction of BMP test duration is possible by predicting the final gas production. The purpose of this thesis work is the development of a new procedure/algorithm to obtain a preliminary estimation of the experimental BMP result by using a linear fractional transformation (LFT) formulation of the original nonlinear model for the identification of its parameters. The efficacy and the efficiency of the algorithm developed will be verified using experimental data of BMP tests performed on different substrates commonly fed to anaerobic digesters.

Reference teachers: Prof. Gianni Ferretti and Arianna Catenacci – gianni.ferretti@polimi.it – arianna.catenacci@polimi.it

P22/23-ENG

Developing educational activities for Control Systems for Smart Agriculture

September 20, 2022December 5, 2022
by admin

This thesis aims at setting up simple experimental devices and experimental activities on logic and modulated control applied to agricultural automation, to support the labs of the course Control Systems for Smart Agriculture.
An example can be a simple automatic irrigation system, including one or more tanks with level control, one or more pumps with flow control, a moisture sensor, etc.

The work includes the set up of the experimental device and of an experimentally validated simulator, the design of the control system, its implementation on an Arduino board, and its validation in simulation and with suitable experiments.

This work is suitable for students with an engineering or non engineering background, and for a single student or for a group of two students.

Reference teacher: Prof. Luca Bascetta – luca.bascetta@polimi.it

P22/23-ENG

Modelling an off-road robot for agricultural applications

September 20, 2022December 5, 2022
by admin

Every time you have to design a control system, you have to first set up a model of the plant, that allows to study its properties, design the controller, and perform a preliminary validation, in simulation, of the control system.
In the case of agricultural robots/machines, the model can be used to test different mechanical designs, as well.

This thesis aims at developing the model of an Agilex robot, suitable for off-road tasks, including the models of terrains with different mechanical characteristics, using Modelica language.
The simulator is then validated comparing the behaviour of the simulated robot with the behaviour of a real platform.

This work is suitable for students with an engineering or non engineering background, and for a single student or for a group of two students.

Reference teacher: Prof. Luca Bascetta – luca.bascetta@polimi.it

P22/23-ENG

Modeling the interaction of a robot with the soil…

August 20, 2022May 21, 2024
by admin

To design a control system, it is first necessary to develop the model of the system to be controlled. This model is useful for studying the system’s properties, designing the controller, and performing a preliminary validation of the control system through simulation.

In the case of robots and agricultural machines, the model can also be used to verify the results of different approaches to the mechanical design of the system. To achieve this and validate the control system in a realistic context, an accurate soil model is also required.

This thesis aims to extend the model of an Agilex tracked robot, developed in a previous thesis, by including an accurate soil model. The model will then be experimentally validated by comparing the behavior of the simulated robot with that of a real platform.

This work is suitable for students from both engineering and non-engineering backgrounds and can be undertaken by either a single student or a group of two students.

Faculty advisor: Prof. Gianni Ferretti – gianni.ferretti@polimi.it