AUTOMOTIVE

See Titles
Title: Realisation of a measuring instrument for estimating the moisture content of straw by means of frequency analysis of capacity/impedance.

Short Abstract
This project aims to estimate the moisture content of straw by analysing how the capacitance and impedance of the material varies with frequency. To do this, Analog Devices’ ADuCM350 board will be used. The work involves setting up a measurement system, collecting data on straw samples with different moisture levels and processing the results to find a reliable correlation. The project combines aspects of analogue electronics, measurement, and data analysis, with practical applications in agricultural monitoring.

Title: Development of a multimodal mixed signal system for biosignal acquisition based on photoplethysmography and skin conductance.


Short Abstract
This project involves the development of a multi-mode mixed-signal system for the acquisition of biological signals, in particular photoplethysmography (PPG) and skin conductance (EDA). The student will design the analogue and digital systems required to acquire, condition and digitise these signals. The system will integrate multiple sensing modes to simultaneously capture cardiovascular and electrodermal activity, enabling richer physiological analysis. The main tasks include hardware design, signal processing and system validation.

Title: Sensor fusion algorithms for non-intrusive and robust measurements of driver biosignals.

Short Abstract
This project focuses on developing sensor fusion algorithms to estimate driver biosignals in a non-intrusive and robust way. The student will work with data from multiple sensors, such as EDA PPG , ECG, pressure sensors on the steering wheel, seat position sensors, and others if needed realized by the student. The goal is to combine these signals to extract meaningful physiological and behavioral information about the driver. The project involves signal preprocessing, feature extraction, and the design of fusion strategies to improve reliability and accuracy and electronic design.

Title: LiDAR-based object detection in an auotomotive environment using Machine Learning/Deep Learning models in Python and MATLAB.

Short Abstract
This thesis proposes a point cloud LiDAR-based object detection algorithm optimised for execution on microcontrollers with FPGAs and Carfield architecture. The method uses lightweight 3D clustering and spatial preprocessing techniques to ensure efficiency and accuracy on resource-limited hardware. The implementation is designed to operate in real time with low power consumption and high parallelisation.

IOT

See Titles
Title: Analysis and Optimization of LoRa protocols for IoT measurement systems, with practical implementation on Raspberry.

Short Abstract
With the advent of the Industry 4.0 paradigm, IoT measurement systems became of fundamental importance, to smarly (and even wirelessly) connect sensors and collect measures. Low Power Wide Area Networks (LP-WANs) are surely valid protocols to handle time-critical industrial traffic. The objective of this work is to develop and optimize a LoRa communication network, specifically designed for industrial measurements. In particular, LoRa modules will be controlled by using raspberry devices, where firmware and software must be implemented (exploiting C and Python languages), pointing to increase the measurement accuracy. Suitable Machine Learning techniques will be also exploited, to optimize the network configuration. 

Title: Vision Based Measurements Systems: Exploiting Machine Learning Techniques to Derive Accurate Measures from Ophthalmic Images.

Short Abstract
The application of Machine Vision to the instrumentation and measurement field, namely Vision–Based Measurement (VBM) systems, is nowadays a hot research topic. In the OptoLab laboratory, several ophthalmic instruments are going to be developed, where Artificial Intelligence (in particular Machine Learning Techniques) are going to be exploited. In particular, between others, CNN-based Neural Networks and Generative Adversarial Networks (GANs) may enhance the functionalities of our VBM systems.

Title: Development of a Control System for Light Automotive Applications.

Short Abstract
The main objective of the activity is to develop a hardware controller able to control small internal combustion engines, two stroke or 4 stroke, fueled with gasoline. In order to achieve the best control performances, the controller has to read several analog inputs: ambient pressure, ambient temperature, exhaust gas temperature, throttle position sensor, phonic wheel. The controller has to be designed by exploiting Simulink, because of the flexibility in code generation. The second phase of the work consists in the low level programming of a microcontroller. The final objective is to build up the board prototype (PCB) and to test it in a real engine.

Title: Smart and IoT based measurement systems in the Industry 4.0 era: exploiting Time Sensitive Networking (TSN) protocols to increase measurement accuracy. 

Short Abstract
Sensor networks, in the current industrial scenario, must exploit high-performing hybrid wired and wireless architectures, pointing to the so-called Industrial Internet of Things (I-IoT).  In this context, the exploitation of TSN protocols is an attractive research challenge. Furthermore, as wide experimental testbeds are costly and difficult to implement, the development of realistic simulation tools is of fundamental importance. This work aims at the implementation of a suitable, realistic simulation tool of a distributed measurement system exploiting TSN protocols, by exploiting the Omnet++ framework and C++ / python languages. 

Candidate skills: MATLAB, PCB design, Electronics Design, Microcontroller Programming, Python.