If you wish to propose your own topic in a related field, please feel free to do so and contact us!

Open theses topics for the 2021/2022 study year

List of Supervisors


Making #smarthome smarter with #fog – #cloud computing platform

(Chinmaya Kumar Dehury, chinmaya.dehury@ut.ee)

1.  Monitoring and controlling smart home appliances using IoT devices

This mainly focuses on controlling the smart home appliances to save energy consumption and ultimately reduces the cost in the smart home environment. For this, a student needs to study the pattern of the power usage of each device that is equipped with the IoT device. Based on the pattern, the designed algorithm needs to predict whether to switch off or on the home appliances.

The student needs to use the existing machine learning tools to analyze the power usage pattern.

2.  Reinforcement learning in cloud resource distribution (chinmaya.dehury@ut.ee)

Reinforcement learning is one of three ML paradigms. Here a software agent takes actions by understanding the environment and its experience. For example, finding a path from one location to other, solving a knight-prince problem, etc. There are several frameworks to address different kinds of problem. In this topic, we will study different RL frameworks and will follow the RL approach in order to distribute the cloud resource among different services/users.

  • Understanding the fundamental concept of Reinforcement Learning and Cloud resource distribution
  • Survey of different RL frameworks (such as OpenAI Gym, DeepMind Lab, Amazon SageMaker RL, Dopamine, etc.)
  • Apply the RL approach to distribute the cloud resources among services/users

3.  1-Shot integration of IoT, Fog, and Cloud (chinmaya.dehury@ut.ee)

Single window system to configure and deploy data flow processors, execution stacks such as Function as Service(FaaS) at Fog and Cloud environment. The main objective of the project is to develop a toolset that can configure the fog environment with cloud connection along with necessary softwares required for data driven execution using various commands. 

This is more of a software package oriented thesis topic where the user needs to be involved in further development of an existing project.

4.  AI-powered cloud usage controlling system in smart home environment (chinmaya.dehury@ut.ee)

Master Level
In a smart home environment, we are using cloud computing for powerful processing and storage of big data that are generated from IoT devices. Further we are also using Fog to handle the data in a more real time manner.

This thesis topic mainly focuses on efficient utilization of cloud and fog. An AI model needs to be developed to decide when to use fog computing and when to use cloud computing.

5.  (Already Taken) Pipeline approach for handling data in hybrid cloud

The student will focus on development of the data pipelines using TOSCA language for Google, Amazon cloud. The data pipeline should also be compatible with serverless platforms. Some TOSCA-based data pipelines are developed for private cloud, such as University’s OpenStack cloud. Apache Nifi can be used as the underlined technology.

6.  (Already Taken) Predicting Cloud Resource demands.

As we know, most of the frequently used apps such as Instagram, Twitter, Spotify, etc are deployed on cloud environment. Sometimes the usage of such applications is very high and sometimes the usage is very low. But can we predict how heavy an app will be used in the next few hours? In short, what would be the future demand for a cloud-based service? This is the question, we will answer in this topic.

  • Find out how the cloud resources are allocated to an app/service.
  • Gather the dataset related to the resource usage of different cloud-based applications
  • Apply AI tools to predict and verify the result using the dataset.

7.  CloudTraceBucket: Visualization and Management Platform for Cloud traces

Master Level, Bachelor Level
In this topic, the student will be primarily responsible for collecting publicly available cloud traces, including the behavior of virtual machines, serverless platforms, workloads, and users’ requests. Several cloud data sets are available from Google, Yahoo, Alibaba, Facebook, and other cloud providers. Nowadays, it is hard for any cloud researcher to find a single window for all cloud traces. This is what this topic will address. The CloudTraceBucket will provide the research to visualizes and download the data based on their requirements/conditions. The student will make use of the existing tools for data storage and retrieval.

8.  (Already Taken) Predicting location-based green energy availability in smart building

Master Level
Green energy mainly refers to solar-powered energy. Research question a student needs to address “what amount of solar-powered energy will we get tomorrow or in next 7days, in current and coming session/”. While doing so, the student may need to perform tasks, such as gathering data, preparing data about the cloud coverage, weather, historical performance of solar panels etc.

9.  Blockchain in Edge computing

Master Level
In edge computing, the intelligence is embedded onto the edge devices. For large data processing capacity and in-depth analysis, generally, the data is sent to the fog environment and then the cloud servers, which is far from the sensor’s location. This thesis talks about integrating blockchain for secure processing and transmission of the data from the sensor to the cloud.

Here, the student needs to implement the existing blockchain technologies (e.g. IBM IoT blockchain) at least in a lab environment that mimics the edge-cloud environment. The goal here is to store the sensor-generated data securely in fog and cloud environments using blockchain technology. The student may need to use the AWS cloud or UT’s OpenStack private cloud for a cloud environment.

10.  A systematic survey of blockchain in smart city

Master level, Bachelor Level
Blockchain, a distributed ledger technology, can be used for securing and making the data/event immutable, storage, process, and transmission of transactions/data. This thesis aims to investigate the application of blockchain in smart city areas, such as healthcare, agriculture, supply chain and logistics, mobility, transportation, marriage, event management, etc.

The student needs to investigate where and how blockchain technology is used to address/handle several problems in the above-mentioned smart city areas. The student may focus on particular areas. This thesis would mostly be a theoretical work based on an extensive literature survey.

11.  Clustered Edge Intelligence

Master level
You know about clustering the computing nodes, servers, containers or clustering the data, or the IoT devices. Several clustering algorithms are designed and developed for data analysis.  But this thesis is talking about clustering the edge intelligence.

Clustered Edge Intelligence (CEI) works with edge computing concepts. In edge computing, the intelligence is embedded onto the edge devices. For efficient usage of the resource and power constraint IoT devices, different clustering mechanisms are developed. The major limitation with such an approach is that it is mainly focused on the device level. This is where CEI differs from conventional edge intelligence. In CEI, the devices are clustered based on the specific task and the implemented intelligence. As a result, each cluster may consist of several IoT devices irrespective of their type, location, and properties.

This thesis focuses on surveying the existing research work towards CEI and designing a framework for the same.

12.  Smart home related more topics…

Analyzing energy leakage
Predicting energy usage 
Predicting CO2 level
AIR quality profiling   
…..
Above topics probably will use the Delta resources.

Go to top


Serverless, IoT , Edge and Fog Computing

(Shivananda Poojara, poojara@ut.ee Delta R3033)

1. Predictive maintainance of SD Cards in IoT devices

The memory card faults or persistent failureness is major problem in IoT devices. There are several factors that implicitly effect the failureness of teh SD cars such as deployment of devcies in harsh envrionemnts, development of bad blocks and malware attacks, etc,. So an aim of theses is to setup IoT devices test bed using RPIs, try to collect failure data by emulating the scenarios and design solutions for prediction of the system using machine learning algorithms.

2.  Autoscaling of serverless data pipelines in fog environments fog/edge computing environments.

 In large scale IoT deployments, handling the fast movement of data is crucial. Most of the data movements are triggered by events with uncertainty. This fast inflow of events carrying data, needs to be handled efficiently to enact the QoS requirements.  This can be achieved using various data flow engines such as Apache Nifi and stateless serverless functions. An aim of the given topic is to design novel auto scaling techniques to handle uneven requests of  serverless functions to minimize the service time, latency and other QoS expectations. Scalability can be achieved within the cluster of fog nodes.

3. Usage based inusrance using IoT

Insurance policies historically have been based mostly on how much you drive, but with advanced telemetry and sensor data, it is possible to incorporate actual driving behaviors into insurance risk models. These behaviors include acceleration/deceleration, speed compared to speed limits, and types of driving, such as commuting on freeway compared to commuting on surface streets. Insurers base policies on observed driving behaviour, which means that safe drivers can be rewarded with lower insurance premiums. This can be achieved using IoT technology. So an aim of the theses is to look in to IoT devices and its ecosystem for usage based inusrance sector

4. COSCO: Container Orchestration Using Co-Simulation for Fog Computing Environments

Aim of this theses is to explore COSCO(a fog computing tool) and understand the diffrent container orchestartion algorithms in this tool and propose solution to address the challenges faced in data oriented workflow based deployments in edge fog applications. The COSCO built with efficient machine learning algorithms for containr placement with various QoS parameters such as energy, latenacy etc,.

5. Investigation of various managed serverless data pipelines

The goal of the topic is to investigate and benchmark functional and non functional metrics for serverless data pipeline mechanisms provided by Google cloud, Azure cloud and AWS cloud services. 

Go to top


Internet of Things topics

Cataloguing the Smart City and Building data (B, M) (Pelle Jakovits)

Nowadays it is quite common for new buildings to have hundreds or even thousands of sensors that generate data. The same can be said for cities, where data about different systems, devices, vehicles and people are being sensed and collected. Often this data is stored in a technical manner that is the most convenient for the actual hardware devices and networks that are used, but which makes it difficult to understand for humans and complicated to reuse in other applications. The goal of this topic is to investigate how to build both human- and machine-understandable data models for cataloguing the data that is being collected in Smart Buildings (E.g. Delta building) and Smart Cities (e.g. Tartu), with the aim to simplify the understanding and reusability of the data.

Synthetic IoT data generator for large scale IoT Device simulation (M) (Pelle Jakovits)

This topic is related to the Cyber defence simulation of Internet of Things and Mobile Networks in the Cyber Range project.
The student should evaluate extending open source tool for generating real-like IoT data based on existing captured data traces and “play it back” to simulate real data in an IoT network. The goal is to design and create a solution which processes existing data traces and can generate similar behaving data stream with high-volume and frequency. It should also support customizing the generated data stream, including volume and frequency, structure (like ratios between the different types of sub-streams), randomizing certain fields of records.

IoT data analytics for real-time visitor count estimation in the DELTA building (B, M) (Pelle Jakovits)

The Delta Building is a new building to house the Institute of Computer Science. Its construction was finished in 2020. A number of different modern sensors were placed in the building or can be placed in the future. The Computer Graphics and Virtual Reality lab’s students have created a real-time visualization of the people and activities inside the building. For that purpose, there is a desire to know how many people occupy each room (including the hallways) at any given moment. The goal of this topic is to study the state-of-the-art of sensor analytics or image processing (or fusion) and to create a usable approach for real-time visitor count estimation in lecture rooms.

Reliability and performance of industrial IoT platforms (B, M) (Pelle Jakovits)

The goal of this topic is to analyse the reliability and performance of IoT platforms (e.g open-source IoT platforms, cloud services like Cumulocity, Amazon IoT, IBM IoT) focusing on solutions that support large use cases (e.g not home automation but rather Smart City, devices deployed over large geographical locations). The main aspects to focus on are the performance, stability and the scope of features (e.g device integration, integration with external services, data processing and analytics, extensibility). The main questions to answer are:

  • Which of the solutions in the market are the most suitable for supporting large Smart City use cases.
  • How easy is to integrate new IoT devices.
  • Are open source IoT frameworks mature enough for production systems and are they feasible alternatives to cloud-managed subscription-based IoT platforms.

Universal Home hub (B) (Pelle Jakovits, Jakob Mass)

The goal of this topic is to investigate the feasibility and cost of building universal home hubs in comparison to depending on commercial brand home hubs. Each different home automation brand (e.g Philips or Fibaro) comes with its own hub, which makes integrating devices from different brands a complicated approach and may require deploying many different hubs in the same rooms, cluttering the space. Some sub-topics to focus on would be: home automation possibilities on open platforms, cloud Home Robots be used as home automation hubs.

Wireless vs wired home automation (B) (Pelle Jakovits, Jakob Mass)

Most new off-the-shelf Home automation IoT devices use Wireless connections, while typical building automation systems utilize wired connections. The goal of this topic is to investigate what are the disadvantages and advantages of wireless home automation devices, investigate their inter-connectivity, protocols, security, etc. One of the main research questions to answer would be: how many devices can a standard size homes/rooms support before the wireless networks quality starts degrading and what parameters can improve or degrade the quality and performance of such networks.

Go to top

Large Scale Data Processing

Stream data processing on resource constrained devices (B/M) (Pelle Jakovits)

With the ever increasing amount of data that needs to be collected from IoT data sources, it becomes more and more expensive to simply stream all the data to a cloud-side data processing platform. Depending on specific scenarios, it may be beneficial to (pre-)process the data as close to its source as possible. However, there are limitations on how powerful computing resources are available near the data sources. The goal of this thesis is to evaluate existing solutions for streaming data processing which allow performing part of the data processing nearer to the source, give an overview of their usability, advantages and disadvantages and analyse their effectiveness in comparison to more classical stream data processing frameworks such as Apache Spark or Storm.

Distributed Serverless Data Processing in IoT networks (M)  (Pelle Jakovits)

The goal of this topic is to study how efficiently Serverless technologies can be utilized to process data streams in multi layer (Fog computing)  IoT networks in a distributed manner and compare the efficiency, reliability and security of this approach in comparison to the typical Cloud centric data processing.

Go to top


Mobile Applications & Mobile Development

Automated Homework Grader for Android applications (Jakob Mass)

The Mobile Computing & Internet of Things course involves several assignments where students develop an Android application according to a set of requirements. Most of the functionality can usually be validated by means of testing interactions with the UI. The validation and grading could potentially be automated using UI testing tools such as Espresso or Roboelectric. In this thesis, the student should design & implement an automated grading system for Android for use by students and course instructors. The system should allow students to submit compiled Android apps and which give students feedback based on their submission. Further, the different submissions and results should be so that course organizers can keep track which students have successfully submitted which tasks.

Smart Home

Extending rule-based automation in the Home Assistant smart home platform (BSc)

Home Assistant is an open-source software platform which integrates a large selection of commercially available and also DIY IoT devices, supports manual and automated control of the devices in a smart home through UI dashboards, a rule-based automation system and more. This thesis will analyse Home Assistants existing automation system, understand its limits and based on the findings, propose an extension or replacement for it. One such replacement could potentially be based on a logic-based language such as Datalog, allowing more declarative, logical predicate-based rules and automation which are more flexible than a “if-this-then-that” based system. Alternatively, an improvement could be to introduce a system which detect common patterns of the smart home actions and propose to convert common habits into rules for the user. E.g. if the user always switches off all the lights in the living room after the TV is turned on on a weekday, the system could detect this pattern and recommend to automate it.