Masters and Bachelors assignments

The Pervasive Systems group continuously investigates new and exciting research areas and there are a lot of opportunities for Masters and bachelors students to explore new interests or consolidate foundations. Have a look below for some ideas, all the projects can be tailored for either Masters or Bachelors so contact us for more information and also if you have your own proposal for a project that fits Pervasive Systems.

Available assignments

The assignments have been provisionally divided into categories to make reading this page easier. In general, each category represents the application domain in which each research project can be carried out. All projects have a small header at the front specifying if the main work deals with sensing, data science, or networking. The interested students are encouraged to contact the protective supervisors to see how a topic can be adapted.

[S = Sensing,   N = Networking,   D = Data Analytics and Machine Learning] 

Humans and Health

• [SD] Compressed convolutional neural networks for the classification of electrocardiogram signals on embedded devices
• [DS] FCMW Radar: Stay home and listen to your heart!
• [DS] FCMW Radar: Check your pet's heart rate together with your heart rate!
• [DS] FCMW Radar: How you doing buddy? Track and monitor your pet.
• [D] Predicting Subtle Human Behaviour 
• [D] Weighted Convolutional Neural Networks Rare Electrocardiogram Signals Detection for Real-Time Heart Monitoring
• [D] Listen up: Unobtrusive voice sensing for elderly care in agitation
• [SD] Social Robots Are Here To Help
• [SND] Sensor data pre-processing & compression for BLE transmission
• [D] Smartphone's Context Detection using Transformer
• [D] Human/ Animal Activity Recognition Data Analysis
• [SD] Activity Recognition Using Two Radars
• [SD] Gesture/ Posture Recognition Using Spectrogram
• [DN] WiFi vs. Sensors: Face-off between physical sensors and wireless signals in human activity recognition
• [DN] WiPi: Contactless sensing using a Raspberry Pi and NEXMON

The Natural World: animals, plants and biodiversity

• [D] Luistervinq 1 – Activity recognition in nature using sound and AI
• [D] Luistervinq 2 – Activity recognition in nature using sound and AI
• [SND] ElasmoPower - Remote Fish Monitoring using Sensor Devices
• [D] Smart Labeling of Animal Activity Data
• [D] Use Everything you got
• [ND] Enhancing Underground Ad-Hoc Network Performance Under Dynamic Soil Conditions
• [SD]  Hoof balance and growth monitoring 
• [D] IMU Generator with End-To-End Learning for IMU Data Generation from Videos
• [SD] Play with your pet: Analyse your pets activities or vital signs

Wireless networks of the future

• [N] Do Horoscopes Work for Factories? Performance Comparison of Wireless Channel Prediction Algorithms within an Indoor Factory Floor
• [N] Are Horoscopes beneficial? Sensitivity Analysis of Predicting Wireless Channel in multiple environments
• [N] Mobile Ad-hoc Routing in MaritimeManet
• [N] Design and implementation of an improved MaritimeManet node 
• [N] Revitalizing the wireless network infrastructure for MaritimeManet
• [N] Distributed Neighbourhood Discovery in MaritimeManet
• [N] Reliable low-latency networking: many radios, many paths
• [N] Hierarchical and distributed control of wireless networks
• [N] Build a clever Wi-Fi access point
• [ND] Bringing intelligence to Wireless Sensor Nodes
• [DN] Edge AI for a better wireless network
• [DN] One AI to rule them all (the wireless networks)
• [N] Connect: Ad-Hoc LoRa network in extremely lossy environments 
• [ND] Fixing the Wi-Fi in the future?: Predictive Forward Error Correction
• [ND] Finding the Truth: Probabilistic Graphical Models for cross-layer inference
• [SD] Securing your things: Find the anomaly in the IoT
• [SN] CommuniFi: Data transmission during device-free sensing of human activities
• [DN] VariFi: Variable WiFi data rates during device-free sensing
• [N] Bringing Control to Wi-Fi: Dynamic Network Slicing in your Home Router
• [N] ML-enabled Channel Assignment and Power Control in Wi-Fi Networks
• [N] User Mobility Management in Software-Defined Wi-Fi Networks

Localization and Navigation

• [D] Smartphone-Based Indoor Localization using Compressive Transformer
• [D] Transfer Learning for Indoor Localization with Smartphones
• [D] Convolutional Algorithms for Real-Time Data Processing with Time Series on Smartphones for Navigation and Tracking
• [D] Compressed generative adversarial networks for real-time feature extraction for fingerprinting localization with smartphones
• [SD] Am I indoors or outdoors?: Modeling location of smartphones

Logistics and transport

• [D] Federated Learning in Smart-Bikes
• [SD] Road Safety in Smart-Bikes
• [D] A tool for the systematic review of multimodal transport and logistics simulation techniques
• [D] Building a simulation tool based on Cloud-Fog-Edge Computing in multi-modal logistics
• [D] Open IoT data-based visualization tool for transport and logistics movements

Smart infrastructure and predictive maintenance

• [D] From point cloud to CAD model
• [D] Embedded Machine Learning for Real-Time Flow Rate Measurement
• [D] Lightweight image-based key point tracking for real-time bridge monitoring with smartphones
• [D] Towards Simulation Tools for Federated Learning
• [D] Automatic detection and estimation of the area of buildings & pools
• [D] Real-time assistance of endangered people in forest fires
• [D] An autonomous approach to generate the land use land cover classification map from the satellite images

Theses in Industry

The Pervasive Systems group collaborates very actively with many industry and research partners. Below is a list of interesting industry-lead research topics for the students interested in an experience in the industry sector.

• [DN] Attacks against AI/ML-based systems in communication networks @ TNO
• Towards a new process for fieldwork on soil and subsoil with SAXION
• Automation and digitalisation for soil and subsoil fieldwork with SAXION

Group Assignments

MaritimeManet – Mobile Ad-hoc Networking at Sea

Previous Projects

These projects have already been done so they are not available anymore but they can give inspiration! If you see something interesting here, get in touch with the supervisor, a new, similar topic can be usually arranged.

• [DS] Buzz Buzz: Monitoring insects without touching them
• [D] Deep Unsupervised Representation Learning for Animal Activity Recognition
• [D] Which Came First? The AI or the Data?
• [D] Horse Activity Recognition Pipeline
• [D] Let There be IMU data 1
• [D] Let There be IMU data 2