Awareness and insight characterise DCU final year projects
11 August 2020 | 0
Innovation, like art, absorbs its influences from what is happening in the world around us.
With this in mind, there is no surprise that the 2020 final year projects from Dublin City University‘s faculty of Engineering and Computing address such wide ranging topics as remote autonomous vehicles, user experience in virtual reality (VR) applications, smart city applications, deep learning assessment of natural disaster damage, and more.
The sheer breadth of imagination and vision is inspiring from the various classes of newly minted engineers.
While TechPro was primarily interested in the computing projects, diesel hybrid aerospace engines, energy harvesting analysis and collagen-based bone repair packs were hard to ignore.
Alanna Carton and Ronan Barker cooperated on a combined project to examine the possibilities for a Remote or Autonomous Mobile Platform (RAMP).
With applications in the likes of search and rescue, as well as remote inspection and monitoring, the project resulted in a small autonomous vehicle that employs image processing to detect objects via a trained neural network.
Two cameras are mounted on the unit, one to relay day and night vision images to the controller, and another for object detection and logging. It can map its surrounding area using GPS coordinates, altitude, temperature and pressure while relaying this and gyro data to the user in both autonomous and manual modes.
The Internet of Things was a key element of Andrei George Rosu’s efforts to improve VR user experience using beacons. Rosu’s investigates how IoT Bluetooth Low Energy (BLE) Beacons could be used to improve the user experience in VR applications by taking advantage of their continuous data transmission capability.
Data can be received by other Bluetooth devices, such as mobile phones or Raspberry Pis, when within range of the beacons. The data can then be sent to a VR application over the network and can be used to trigger specific events within the application.
Some lecturers have a sixth sense for when they are losing the attention of their class. For those lacking in such empathic skills, David Early investigated a system of emotion detection for lecturers.
Early developed a web application to enable lecturers to monitor audience emotion throughout a lecture, using face detection, emotion classification and speech recognition to provide visual feedback in the form of graphs.
Audience images are collected using the lecturer’s mobile phone and are processed using REST services, making the application transferable to alternatives to university settings.
The goal of the project was to provide lecturers with “unbiased audience emotion information”, thereby enhancing student learning.
Eoin Clayton’s Smart-City is a landmark and building recognition mobile application that aims to remove the manual process of searching for information about a landmark. It achieves this using a pre-trained Convolutional Neural Network independently hosted on a server. The app identifies the building by a photo submitted by the user and displays the name and related facts about the landmark.
The environment, energy and resources figured heavily in the project topics. Eoin Woodward’s project examined whether acoustic sources could yield usable energy, exploring the potential of piezoelectric materials, while Shane Tuite used deep learning-based computer vision to undertake natural disaster damage assessment from satellite images.
While many of these topics are visionary and could have far reaching applications, the eminently practical also featured. Kevin McGonigle developed comet – an intuitive, powerful and comprehensive code metrics and analysis tool.
Using a Python-based parser built on the ANTLR parser generator, comet produces insightful and valuable performance and structural metrics and models for submitted codebases. Requests are serviced by a Django REST API with abstract responses, designed to be useful in the context of comet’s sleek React.js web application, or as an integration with a deployment pipeline to provide a means of implementing metric-based checkpoints into the CI/CD cycle.
Amidst the current public health emergency, and the potential long-term effects for those recovering from Covid-19, Amelia McGuinness investigated the use of 3D printed materials in wearable technologies for breathing rate measurement.
McGuinness examined how conductive inks can be used in the manufacture of wearable devices to optimise their electrical properties for use as sensors.
While this year’s project showcase was a virtual affair, the application of practical learning to such a wide array of topics and issues is the best demonstration of the potential for these students to go out into the professional world and make valid contributions for the future.
Addressing the observers of the digital showcase, Prof Lisa Looney, dean, Faculty of Engineering and Computing, DCU, said: “We are confident that the students you ‘meet’ here will go on to have a strong impact on the various industries and sectors that you represent. We look forward to marking the success of these students in the future.”