48440 research outputs found
Sort by
Using Machine Learning to Predict Network Performance of a Micro-Structured Gas Separator Using the Soret Effect
No full textHydrogen is a promising clean energy source, emitting no CO2 during use. However, production methods rely on fossil fuels, contributing to carbon emissions. To combat this, the Soret Team at the Shibaura Institute of Technology in Tokyo is investigating hydrogen separation using a thermophoretic microstructured separator device. Alternatively, this project simulates using multiple smaller devices in a network structure, employing a Random Cut Forest machine-learning regression model to predict separation. A custom GUI was developed to visualize and interact with this network. Results showed potential benefits in scalability. Future work involves replacing synthetic datasets with experimental data. This project advances clean energy and highlights the benefits of interdisciplinary research
Interchanges along Hong Kong’s waterfront – Where transport and place making meets
No full textHong Kong is a highly dense city built around Victoria Harbour. As dense developments have sprouted throughout its history, the city’s need for reliable, efficient transit has become increasingly crucial. Many hubs have emerged along its iconic waterfront, integrating a multitude of transit options. Over the course of this project, we have observed, analyzed, and proposed improvements to the waterfront interchanges along Hong Kong’s Victoria Harbour; seeking to transform them into not only better transit hubs, but also vibrant public spaces. This was accomplished by establishing a set of criteria against which all 12 waterfront interchanges on Victoria Harbour were evaluated. Data collection, research, and analysis led to the development of ideal standards that all waterfront interchanges on Victoria Harbour should strive to meet, and would allow them to excel in all our criteria. Finally, we applied these standards to four key waterfront interchanges, formulating specific recommendations to help them better align with these benchmarks
Development of an educational mobile robot platform
No full textThis project delivered an open-source solution for a low-cost, educational mobile robot platform based on the Raspberry Pi 5 and Dynamixel system, which can be completely assembled using an entry-level 3D printer with polyethylene terephthalate glycol (PETG) material. The robot is modular, allowing users to design their own chassis or actuators. By default, this project provides a 4-wheel differential chassis that enables the robot to climb a slope and still be able to keep its center of rotation across the geometric center of the robot’s body; a two-axis gimbal compatible with a wide range of camera modules; and a SLAM sensor platform equipped with an RPLIDAR C1 time-of-flight (ToF) lidar, with screw holes left for users to fit their Global Navigation Satellite System (GNSS) or Ultra-Wideband (UWB) positioning sensor on it. An Inertial Measurement Unit (IMU) for sensing the robot’s orientation is also equipped with this robot as part of it. The software of this project is developed based on Ubuntu 24.04 LTS and Robot Operating System 2 (ROS2). The software package includes hardware drivers, independent of ROS2, for IMU and DynamixelSDK, which hides the operation of hardware from the user so that even students at an initial stage of study will be able to learn and use them easily. Additionally, the ROS2 drivers implement the functionality of Simultaneous Localization and Mapping (SLAM) tasks with the lidar and IMU and receive manual control input sent by the Sony DualSense controller with a set of intuitive keybindings. For the map generator node of ROS2, a method is added to effectively remove errors and incorrect scanning results in the map that happen when the robot turns. The cost of the robot system is estimated at around 298 for the minimum system. The GitHub link of this project is provided here: https://github.com/vincent-tvincent/MQP2025-educational-mobile-robot/tree/main
Project 3: Strategic Planning for Post-Disaster Logistics
No full textIstanbul is located along a fault line and has a history of earthquakes above level 7. Istanbul is the largest city in Europe, and is continuing to grow, and adapting aid logistics to said growth is a great opportunity. The 2023 Turkey-Syria earthquake's impact at Hattay not only showed areas for improvement in search and rescue teams, but it also showed that aid, communication logistics, and survivor welfare also need to improve. There was a well-documented shortage of tents and heating in the aftermath of the earthquake. Aid or medical teams were also under-supplied in some areas but redundant in others. Türkiye has a variety of aid sources, including government and NGO. There are three organizations that we were particularly interested in. First is AFAD, which is the central disaster agency in Türkiye. They are still a relatively new agency, being formed in 2009, and they have been tasked with many responsibilities, some of which were formerly assigned to military and local groups. Kızilay is a much older institution, and they cooperate with AFAD to a great extent; the buildings where we interviewed both institutions were even next to each other. A less government-aligned NGO that we interviewed was AKUT VAKFI, the independent logistics arm of the search and rescue NGO AKUT. A recurring problem throughout our interviews is trust and communication between groups supplying aid. In interviewing agencies that supply aid in these circumstances, we learned about potential areas for improvement in aid logistics, survivor welfare, and communication logistics
Cognisound – Exploring a Low-Barrier Personalized Listening Support Solution for Auditory Processing Disorder
No full textCognisound is a mobile application prototype designed to support patients with Auditory Processing Disorder (APD), a condition that causes the brain to struggle with interpreting sounds despite normal hearing ability. Existing treatment options are often costly and inaccessible. This project explores the design of a low-barrier, user-centered toolkit aimed at adolescents and adults to provide real-time transcription, noise filtering, journaling, and coping strategies. Through the use of design thinking, user personas, and iterative prototyping, the app is grounded in real user needs. Cognisound offers a practical and customizable approach to aid users in managing their APD symptoms
From Newcomers to Locals: Supporting Interconnectivity with SINGA in Lyon, France
No full textMigrating to Lyon, a major hub for immigration in France, presents numerous challenges, from navigating unfamiliar systems to overcoming social exclusion. To address this, our team collaborated with SINGA Lyon, an organization dedicated to fostering connections between newcomers and locals. From a Participatory Action Research (PAR) stance, we employed a mixed-method approach comprised of surveys, interviews, comparative research, and participant observation to evaluate and enhance two major initiatives: INSPIRE, a biennial storytelling event celebrating newcomer experiences, and SINGA’s upcoming participation in the city-wide Fête des Lumières. Through these efforts, we contributed to SINGA’s mission of promoting inclusivity and reshaping perceptions of migration in Lyon
Innovating Innovate: Assessing and Ameliorating Barriers to Submission in the V&A's National Schools Challenge
No full textIn the UK, design and technology education for students in Key stage 3 has been on the decline. Because of this, programs such as the V&A Innovate National Schools Challenge have had issues with engagement. Through surveys and interviews with educators and staff, as well as case studies of similar design challenges in the UK, we have put together a comprehensive list of recommendations to enhance the submission rate of V&A Innovate. We produced a sample lesson plan, reorganized the existing materials, and offered other recommendations including a page dedicated to “First-time Innovators” and ECT’s
Enhancing Environmental Regeneration via Community Corals in Refshaleøen
No full textCommunity Corals is a regenerative project run by Oleg Koefoed and the Center for Livskraft Og Regeneration, with whom we collaborated to connect community and biodiversity. This project enhanced community engagement, interaction and interest in the corals located in Refshaleøen, a former shipyard, Copenhagen. The team engaged in walkshops, spiritual events that connect people with nature, and developed recommendations to improve the Community Corals experience. We introduced staining corals using moss-friendly dyes, proposed concepts for greenhouse domes to improve moss growth, and recommended adding light or sound designs to the corals. Our final deliverable was demonstrating a staining walkshop as a model for engagement with the corals, transforming them into a full sensory experience
Stock Market Simulation 2511
No full textThis project was a five-week stock market simulation using two contrasting trading methods. The main goal of this project was to learn about the American stock market and how to properly trade, as well as identify the key differences between investment and speculation trading. The project used a custom excel-based simulation method which referenced price data from up-to-date online sources. The two trading methods tested were a conventional, active and informed trading method, and a data-driven method which used Google Trends data and a custom decision matrix to make decisions. Both methods traded the same 5 stocks, chosen for their relative stability and growth potential, and a starting budget of $99,983.95. Each began with an identical portfolio with the funds evenly split between all stocks. By the end of the 5-week simulation period, the informed strategy returned a 10.05% increase in portfolio value, while the data-driven method returned 9.84%. The results indicated that a variety of trade decisions are capable of producing strong, returns greater than the S&P 500 (which had 4.37% return), so long as the stocks within the portfolio are selected carefully. The process of participating in this simulation provided a deeper understanding of the stock market and may aid in future investment endeavors
Supporting Professionalism Within IMGD
No full textOn this IQP, the team researched methods for teaching professionalism and applied these methods in order to improve the professionalism of as many Interactive Media and Game Development students as possible at Worcester Polytechnic Institute. These methods involved organizing events that motivated students to create portfolio content, communicate their ideas and skills, network with industry professionals, interact with other roles in their profession, and practice teamwork