PhD, Internship, Visiting Researcher
I am always eager to hear from prospective PhD students with an interest and strong background in Sensors, Systems and Machine Learning, IoT and Distributed Systems. Email me if you are interested.
Title: Federated Learning Algorithms for anomaly detection and predictions in IoT
Abstract: Anomalies detection in distributed networks and cyber-physical systems (CPS) is a challenging task and require multi-objective optimisation. In recent years, Federated machine learning is used for privacy-preserved machine learning. In Federated Learning many decentralised distributed computing nodes (edge or servers) train a local model on local data and then local model are combined centrally to generate a global model. There are different frameworks and algorithms to implement a Federated Learning system. In this project, the aim to design and implement an efficient Federated Machine Learning Algorithm for anomaly detection and prediction in IoT. The proposed work can be breakdown into three tasks: (1) Collect and analyse real-time anomaly data from IoT devices that are publically available. By using custom python packages and tsfresh package, the essential features from the IoT data can be extracted to define anomalies and data outliers. (2) The data can be used to train on-device models and federated on-device model training. (3) Deploy and evaluate the performance of various customised models on state-of-the-art IoT development kits
Prerequisites: The student is expected to have a good understanding of algorithms, and some basic hardware, such as adafruit boards(https://www.adafruit.com/category/957) or raspberry pi (https://www.raspberrypi.org/products/raspberry-pi-4-model-b/) and good programming skills (Python, Java, or C++).
The Start date and duration of the internship is flexible, however, it should be between 8-10Weeks.
Project Title: SmartCity: Data negotiability in multi-mode communication networks
As the concept of the smart city is growing from abstract ideas and a lab environment to real deployments, many intelligent sensor-based IoT applications are emerging. Some of these applications use a camera or audio-based sensors in public spaces such as streets, parks, train stations, public squares and stadiums. Many of these deployments are used for a specific purpose and dedicated infrastructure is setup for the applications, for example, City council CCTV (close circuit television) installed in streets and public spaces for security surveillance. Similarly, the InLinkUK free wifi kiosks in public spaces installed by BT provides dedicated free internet service with the partnership of an advertising company. All these examples are independent solutions in the public space. If these resources and infrastructure could be shared, this will create new services and better co-created smart cities and speed up the deployment of new services in an economical way. In order to deploy many of these smart sensor-based applications using shared infrastructure and resources, a well-structured data negotiation framework is required which complies with GDPR data regulations as well as citizen's privacy in public places. To understand this further, we need to investigate the following questions for building the data legibility and negotiability framework.
Q1: Who supports the necessary public IoT infrastructure, and why? Q2: Who will benefit from this infrastructure and what are the ethical, cultural, social, technical and economic barriers in achieving full benefit? Q3: Who has access to the public space citizen data and how this data is going to be used? Q4: What current mechanisms are used to take data usage consent from an end-user and how they are informed by other service users and providers that are involved in the whole ecosystem? To make progress towards this aim, in the current project proposal, you start with a case study example.
Prerequisites: The student is expected to have a good understanding of wireless protocols, and some basic hardware, such as adafruit boards(https://www.adafruit.com/category/957) or raspberry pi (https://www.raspberrypi.org/products/raspberry-pi-4-model-b/) and good programming skills (Python, Java, or C++).
The Start date and duration of the internship is flexible, however, it should be between 8-15 Weeks.
Call for Summer Internships sponsored by Jayne Lawrence (Computer Science Department Manager) and YorRobots. There are 9 internships available at a value of £261 per week, for a period of 8 to 10 weeks, each. The goal of this initiative is to involve students *from any department* at the University of York in robotics research supervised by one or two academics, also from the University of York. We welcome applications *from students* on topics related to the “Robotics” remit of EPSRC.
Requirements for the project:
Deadline for final submission is 29th of May, therefore, you must contact asap to discuss the plan.
Applications will be reviewed by the members of the YorRobots Executive Committee. Successful applicants will be notified by the 8th of June. A final report on the results of the project will be requested in October.