Academic supervisor:

Íñigo Adín

CEIT Division:

ICT Division

Thematic area:

IoT, communications, low power, Bluetooth.

Description and objectives:

This project proposes the evaluation of Bluetooth 5.1 technology in its Long Range mode for use in industrial environments. The aim is to see the possibilities of this longer range version to compete in IoT technologies to be installed in production plants or in remote sensors, with range for the admissible meters in this case.

The aim is to evaluate this technology using evaluation boards and configuring the elements at FW and SW level to measure the signal quality in parameters potentially useful in industrial applications (time of flight, data rate, packet error rate, etc.). To do this, a search for these evaluation boards will be proposed and programming will be required using the admissible interfaces in each case.

Academic supervisor:

Ion Irizar

CEIT Division:

Data Analysis and Information Management

Thematic area:

Artificial Intelligence

Description and objectives:

Natural language processing (NLP) is a very promising technology for facilitating interaction between people and industrial processes. Currently, such interaction takes place through conventional methods based on screens and keyboards. However, in recent years there has been a growing interest in exploring the potential of NLP techniques in industry.

The task of this GFP will be to perform a literature review of the state of the art on NLP-based industrial applications and available free platforms. Once this review has been carried out, the second objective will be to develop an NLP-based algorithm to interact with a virtual water treatment plant.

Academic supervisor:

Ion Irizar

CEIT Division:

Data Analysis and Information Management

Thematic area:

Control

Description and objectives:

In many industrial processes there are variables that, being of great interest for decision making, cannot be measured directly by sensors. However, sometimes these variables can be observed indirectly by relating them to other measurable variables. The Kalman filter is precisely an observer that allows observing non-measurable variables using dynamic mathematical models involving measurable and non-measurable variables.

The task of this GFP will be to use a simulator of a water treatment plant to develop an observer capable of estimating in real time the concentration of ammonium and nitrate in the plant. Matlab/Simulink and/or Python will be used to carry out the project.

Academic supervisor:

Ion Irizar

CEIT Division:

Data Analysis and Information Management

Thematic area:

Artificial Intelligence

Description and objectives:

Classical automatic control is giving way to new, more sophisticated control techniques based on artificial intelligence, such as Reinforcement Learning techniques.

The task of this GFP will be to use a water treatment plant simulator and apply Reinforcement Learning techniques to design an automatic control strategy that adjusts the oxygen level to keep the ammonium level close to a reference setpoint. Matlab/Simulink and/or Python will be used to carry out the project.

Academic supervisor:

Ion Irizar

CEIT Division:

Data Analysis and Information Management

Thematic area:

Artificial Intelligence

Description and objectives:

Many real industrial processes can be mathematically formulated using complex mechanistic models composed of nonlinear differential equations. Although these models are very useful for carrying out design and operation studies, their handicap is their high computational cost which makes them unfeasible for use in real time decision making.

With the advent of Deep Learning techniques, proposals have emerged to reduce the complexity of these models and thus the computational cost. The task of this PFG will be to use the technique known as "physics informed neural networks" to obtain a reduced model of a water treatment plant. The Python environment will be used to carry out the project.

Academic supervisor:

Ion Irizar

CEIT Division:

Data Analysis and Information Management

Thematic area:

Artificial Intelligence

Description and objectives:

Water treatment plants are subject to increasingly demanding operational requirements. It is no longer sufficient to comply with the quality of the treated water, but it is also necessary to do so with the minimum energy consumption. To achieve this, the operators of these processes need to have adequate information to enable them to make better decisions.

The task of this PFG would be to use an already developed WWTP simulator to generate data sets that collect its historical operation. These data sets will then be used to evaluate different Machine Learning classification algorithms with the objective of being able to predict the operational state of the process. The algorithms will be programmed in Python.

Academic supervisor:

Santiago Figueroa Lorenzo

CEIT Division:

ICT Division. Data Analysis and Information Management Group.

Thematic area:

Software development, privacy, code automation, containers.

Description and objectives:

Fides is an open source privacy management platform that enables the enforcement of privacy standards at the code level. Fides tools allow you to tag system privacy features, orchestrate compliance with programmatic rights, and audit personally identifiable information stored across systems and application infrastructures. Fides in turn supports all major privacy regulations (e.g. GDPR, CCPA and LGPD), and standards such as ISO 19944 by default.

This project seeks the automated implementation of the Fides platform in a practical use case applied to an e-commerce store, so that a consistent and versioned definition of the privacy features and resources of this system can be created, being used as part of a CI/CD pipeline to process privacy requests.

Academic supervisor:

Santiago Figueroa Lorenzo

CEIT Division:

ICT Division. Data Analysis and Information Management Group.

Thematic area:

Security, Software development, Authentication, FIDO2, Authorization, Oauth2.

Description and objectives:

The use of passwords as an element of system authentication has weaknesses related to the limited human capacity to retain long strings of characters, while at the same time being susceptible to different types of attacks such as brute force attacks. This has gradually led to the emergence of tools and schemes that complement the exclusive use of passwords. By default, an authentication system must consist of "something we know", e.g. a password or PIN code, "something we possess", e.g. a credit card or RSA token, or "something we are" (biometric authentication), such as the shape of our hand or fingerprint.

This project consists of implementing a biometric authentication system based on the FIDO 2 standard, which involves the design and implementation of a wallet for credential storage that interacts with an authentication and authorization system composed of Identity Providers, Policy Enforcement Points (PEP) Proxies and Authorization Servers.

Academic supervisor:

Santiago Figueroa Lorenzo

CEIT Division:

ICT Division. Data Analysis and Information Management Group.

Thematic area:

Software development, Security.

Description and objectives:

The project seeks to identify, according to the state of the art, tools and methodologies for the development of secure SW, in an agile and efficient way. Based on this study, the implementation of a use case that integrates the selected technologies is sought. Therefore, the following is sought:

• Identify methodologies for secure SW development: code analysis, testing.

• Identify tools for secure and agile SW development: secure SW development in practice with existing tools.

• Relationship with Continuous Integration and Continuous Deployment tools.

• Testing to support the potential of the identified methodologies and tools.

Some of the tools proposed to be used throughout the project will be Jira, Gitlab/Github, Docker, Kubernetes, Trivy, Jest, Testpy.

Academic supervisor:

Santiago Figueroa Lorenzo

CEIT Division:

ICT Division. Data Analysis and Information Management Group.

Thematic area:

5G, Kubernetes, Cloud Computing, Virtualization.

Description and objectives:

Cloud Native adoption in 5G telecom systems has been identified as a good candidate to reduce cost, improve system agility and the role of 5G services. Based on the 3GPP standard, the European Telecommunications Standards Institute (ETSI) has published the NFV reference architecture tailored to Cloud Native environments and to enhance the NFV framework, including, containers, load balancers and other elements as part of the reference architecture.

This work pursues the validation of container technology on the MANO platform hosted at ETSI, in a CN environment, so that the results obtained in the work can help to encourage users and operators to use KNFs and thus taking advantage of container technologies.

Academic supervisor:

Ion Irizar

CEIT Division:

Data Analysis and Information Management

Thematic area:

Artificial Intelligence

Description and objectives:

The detection of rare events in industrial processes is a topic that, with Industry 4.0, has become particularly relevant. Industrial processes are becoming increasingly digitized which brings significant benefits in terms of improved efficiency, but on the other hand makes these systems more vulnerable to cyber-attacks. This is a particularly relevant issue in the case of critical infrastructures such as water treatment facilities.

Academic supervisor:

José Sebastián Gutiérrez Calderón

Department:

Control and Robotics Group of the Department of Mechanical and Materials Engineering at Tecnun

Thematic area:

Automation and Electronic Systems Engineering

Description and objectives:

Industry 4.0 is currently revolutionizing the way companies manufacture, improve and distribute their products. These smart factories are now integrating a large number of sensors and actuators, and with the help of the Internet of Things (IoT), cyber-physical systems and cloud computing, are making it possible to collect and analyze data for decision making.

The objective of this PFG would be the automation of a monitored bench, simulating an industrial equipment through a PLC and obtaining data from sensors and actuators through the Internet of Things for further processing.

Academic supervisor:

Leticia Zamora Cadenas - Iker Aguinaga Hoyos.

CEIT Division:

Information and Communication Technologies. Intelligent Systems for Industry 4.0 Group.

Thematic area:

Telecommunication/Industrial Engineering

Description and objectives:

Indoor location systems have been on the rise in recent years. Whether using radio frequency technologies, inertial sensors or artificial vision systems, the location of objects or people in indoor spaces is a key element in many applications (tracking of parts, access to security areas, tracking of people, augmented reality, etc.).

To determine and evaluate the accuracy of a location system, it is most common to use manual measurement of control points or tests in a controlled environment to determine the accuracy of the system. However, this type of measurement is always subject to measurement errors, human error, and the impossibility of tracking a moving element in real time. Another widespread option, especially when it comes to assessing dynamic accuracy, is to resort to cost-effective systems that allow the creation of ground truth, such as, for example, vision tracking systems. However, it is not always possible to deploy such systems, or the financial means to do so are not always available. Therefore, being able to assess the accuracy of indoor positioning systems at low cost is still a problem that researchers and companies are trying to solve.

Ceit currently has a line of research associated with positioning systems for indoor spaces, in which it works with various companies to provide solutions to their needs. This is why there is a need for a ground truth system that is easy to install and not too expensive.

The task of this GFP would be to develop a ground truth system, using virtual/augmented reality systems, for subsequent use in the evaluation of the accuracy of Ceit's proprietary indoor location system. HTC Vice, Oculus Quest and Hololens 2 hardware are available for the development of this system using the Unity3D programming platform. The candidate should have programming skills in C# or similar languages such as C++ or Java.

Academic supervisor:

Emilio Sánchez Tapia

CEIT Division:

Information and communications technologies. Intelligent Systems for Industry 4.0 Group. Vision and Robotics Subgroup

Thematic area:

Robotics Engineering

Description and objectives:

Industry 4.0 has paved the way for multiple forms of automation that aim to improve productivity and optimise work processes. In this context, the aim is to develop an intelligent mobile manipulator: a new type of robot that integrates the technology of an autonomous mobile robot and a highly efficient collaborative robotic arm capable of performing various operations.

The idea of the project is to develop a robot that can move, detect and avoid obstacles, explore its environment to recognise objects through artificial vision and carry out part handling tasks, being able to interact with operators. With the idea of implementing a digital transformation model, required today in real factory environments, the robots, control elements, sensors and the rest of the on-board elements will be connected to each other through a digital platform to control the process in real time and from anywhere.

CEIT has already developed a first working prototype (see figure below).

The task of this GFP would be the programming under ROS-2 of a sequence of tasks for the robot to interact with a classic robotic cell. The specific case to be developed will be for the robot to go to a repository of parts to be processed, bring them to the cell, wait for them to be processed and take them to another warehouse of parts already classified.

Under this simple task, the concepts of:

• Collaborative mobile robotics
• Machine tending
• Control in force
• Problem of synchronisation of two automatic devices

Programming skills in C/C++, Python or java-script are required.

Academic supervisor:

Diego Borro

CEIT Division:

Intelligent Systems for Industry 4.0

Thematic area:

Telecommunication Systems Engineering

Description and objectives:

Copernicus is the world's largest Earth observation programme to date. The programme is a joint initiative of the European Commission and the European Space Agency (ESA) to build an autonomous Earth observation system. The Copernicus programme relies on a family of satellites called Sentinel, owned by the European Union and developed to meet the needs of the Copernicus services and their users.

Every object on the Earth's surface reflects and absorbs energy in various ways. The spectral signature represents the unique way in which a surface reflects the sun's energy, within the electromagnetic spectrum. In addition, spectral signatures are typically characterised on an X-axis (wavelength) and Y-axis (percent reflectance) graph so that different surfaces have different spectral signatures.

There are currently a multitude of data, tools and software,... to access and process satellite data. The objective of this PFG would be to access multispectral information from a certain area of the planet and process it to obtain certain information such as changes in land cover, the growth of algae in the water, different types of crops, or the amount of urban development in an area. The specific information needed depends on the application and will be defined when the PFG starts.

The student will not start from scratch as a PFM has been defended and a study of the art of all existing technologies and tools has been carried out.

Academic supervisor:

Santiago Figueroa Lorenzo.

CEIT Division:

DAIM.

Thematic area:

Telecommunications/Computer Engineering.

Description:

Device virtualisation is an important focus of attention today. The open source project OpenPLC virtualises Programmable Logic Controllers (PLCs) according to the IEC 61131-3 standard. In the article "OpenPLC: An IEC 61,131-3 compliant open source industrial controller for cyber security research" a comparative study of the OpenPLC project with real PLCs was carried out with regard to the functionality of the OpenPLC project, with satisfactory results. However, there is currently no comparative study considering exclusively network parameters such as throughput and latency. It is precisely this end-of-degree project (PFG) that aims to focus on this analysis. To develop the tests both on OpenPLC and on real PLCs, the "Hardware in the Loop" methodology will be used, which will be carried out by simulating a process in Matlab Simulink which will be connected to the inputs and outputs of the PLCs. Additionally, Modbus clients deployed on the docker tool will be used to carry out the connection tests on the PLCs.

Objectives:

• Connect to each PLC a process deployed on Matlab Simulink as part of the "Hardware in the Loop" methodology.
• Create Modbus clients on docker with the ability to send ad hoc Modbus frames to the process deployed on PLCs based on the pymodbus library.
• Measure latencies and performances of available physical and virtual PLCs.

Academic supervisor:

Santiago Figueroa Lorenzo.

CEIT Division:

DAIM.

Thematic area:

Telecommunications/Computer Engineering.

Description:

Intrusion Detection Systems, e.g. Suricata, are integrated as part of industrial networks where they play a key role in detecting vulnerabilities and threats. These systems generate logs of information which must be transmitted throughout the network for proper storage and subsequent visualisation. However, this transmission and storage of information is also susceptible to cyber-attacks, hence the following project aims to design a system that guarantees the reliable transmission of monitoring logs.

Objectives:

• Design and implement agents that capture, adjust and minimise logs, and then invoke transactions to the decentralised network based on Hyperledger Fabric Blockchain.
• Design and implement the smart contract that will allow the reception of logs and their subsequent storage in an optimal way.
• Determine, from a performance point of view, the best method of log storage: internal vs. external storage with pointer to the blockchain.
• Determine, from a performance point of view, how many logs per second an agent is capable of storing.
• Develop a role-based access control policy that guarantees access to logs for certain users (depending on progress).
• Develop a basic UI for the visualisation of the logs consulted from Hyperledger Fabric Blockchain (depending on the progress).

• Title:
  Internal communication system for large MPVs

• Academic Supervisor:
  Adam Podhorski

• Recommended profile:
  Degree in Telecommunication Systems Engineering
  Degree in Communications Electronics Engineering

• Description and Objectives:
  In large MPVs it is quite difficult to maintain a conversation between passengers in different rows due to a higher noise level than in conventional passenger cars. The aim of this project is to review the ideas for solving this problem proposed so far, to design a multi-channel internal communication system with echo cancellation, to simulate it in Matlab and to realise a prototype based on a C6713DSK DSP board equipped with a multi-channel audio interface and to test it in a car.

• Area/Department Ceit/Tecnun:
  Department of Electronics and Communications - Ceit

• Subject area:
  Telecommunication Systems Engineering / Communications Electronics Engineering.

• Title:
  Diver voice correction.

• Academic Supervisor:
  Adam Podhorski

• Recommended profile:
  Degree in Communications Electronics Engineering
  Degree in Telecommunication Systems Engineering

• Description and Objectives:
  Divers working at great depths breathe a mixture of helium and oxygen (heliox), because at high pressures nitrogen is harmful. However, this mixture causes a change in the voice, which takes on a Donald Duck-like characteristic and becomes practically unintelligible. One of the solutions to this problem is to construct a map in which the spectra of different sounds produced in heliox and in air by the same person are related. The voice produced in heliox is corrected by changing the spectrum of each sound by its corresponding spectrum of normal voice. The aim of this project is to investigate a number of changes and possible improvements in the quality of the corrected speech for an existing Matlab design of such a system.

• Area/Department Ceit/Tecnun:
  Department of Electronics and Communications - Ceit

• Subject area:
  Telecommunication Systems Engineering / Communications Electronics Engineering.

Academic supervisor:

Diego Borro

Department Tecnun/CEIT Division:

CEIT. Intelligent Systems for Industry 4.0 Group.

Thematic area:

Machine vision for industry

Description and objectives:

An application is available that allows testing a multitude of artificial vision algorithms on images. The project will consist of increasing the capabilities of the current tool by adding algorithms, improving the GUI, providing more functionalities, etc.

Academic supervisor:

Diego Borro

Department Tecnun/CEIT Division:

Intelligent Systems for Industry 4.0 Group of CEIT

Thematic area:

Machine vision for industry

Description and objectives:

The project consists of using Deep Learning techniques for object detection and classification. Specifically, the focus will be on detecting parts with and without defects.

Academic supervisor:

Íñigo Adín.

Department Tecnun/CEIT Division:

CEIT. ICT Division

Thematic area:

Positioning and communications

Description and objectives:

This project aims to propose novel techniques to achieve indoor and outdoor positioning by means of signals of opportunity, in combination with GNSS satellite signals (GPS, Galileo, etc.). The current trend is to fuse the GNSS signal with inertial sensors and vision and radar based sensors for precise guidance. However, there are less expensive techniques using signals already present in the spectrum and in the application environment that can be useful. Wifi and Bluetooth beacons can be used as anchors and fingerprinting techniques based on received power can achieve accuracies in the order of a few metres. However, the new 5G, with higher frequency spectra and smaller cells, can be useful for positioning by signal arrival time. There are also constellations of commercial satellites in low orbits that emit signals that can also be used for this purpose.

The aim is to establish the techniques used in each case and to arrive at an estimate of the achievable accuracies.