48440 research outputs found
Sort by
Implementing Good Agricultural Practices Among Thai Mango Farmers
No full textIn this project, we investigated the adoption of Good Agricultural Practices (GAP) among Thai mango farmers. We sought to learn the challenges they are facing and their compliance with the GAP regulations. We collaborated with The Thai Agricultural Innovation Trade Association (TAITA) to encourage Thai farmers to implement GAP certification and to come up with strategies for increasing the adoption rate among Thai farmers. Our research methods included archival research, surveys, and semi-structured interviews with the President of the Thai Mango Association, Thai mango farmers, mango vendors at Simummuang Market, and consumers. We found that many Thai farmers struggle with the GAP regulations, especially the eighth regulation focusing on data collection. We also found GAP certification is not highly prioritized in the domestic market due to the lack of financial incentives for farmers and lack of Thai consumer awareness. Additionally, Thai mango farmers have limited export markets resulting in declining sales and lower income. Our recommendations are a data collection tracker to solve data collection problems where most of the Thai farmers rely on memory rather than taking note and point deduction system to ensure that Thai farmers will obey the GAP standards efficiently. Another recommendation is raising consumer awareness on the importance of GAP and how it benefits them through the campaign and expanding into new markets. This ensures Thai farmers have an opportunity for exporting Thai GAP products. Lastly, we come up with the detection of fake chemicals making sure that the chemicals we used are authentic
Evaluating Educational Materials that Teach the Risks of Prenatal Alcohol Exposure
No full textThe National Organisation for FASD works to spread awareness of Fetal Alcohol Spectrum Disorder (FASD) in the UK. Despite the high prevalence of FASD in the UK, there is little formal education about the risks of prenatal alcohol exposure including FASD. The goal of this project was to evaluate the effectiveness of National FASD’s educational materials on FASD for young adults ages 16-18. We recruited three lecturers in Furthering Education colleges to pilot the lesson. Through the assessments and lecturer interviews, we determined that the lesson was engaging and successful in conveying key concepts. We also found that the lesson was easy to implement and suitable for the age group. Drawing on student and lecturer feedback, we recommend adjustments in the lesson plan
FinTech Project B24 - Fidelity Enterprise Development Group - Early Career Portal (ECP)
No full textFidelity Investments seeks to optimize the conversion rates of getting interns hired as full-time employees to enhance workforce planning. The Early Career Portal (ECP) project focused on developing a tool to analyze manager-specific data, determine conversion success, and optimize the assignment of interns accordingly. The ECP tool used a Mixed Integer Quadratic Programming model, written in Python, to compute the optimal number of interns for each manager and location. This model, utilized through the full stack web application, was informed by data on intern conversion rates residing in Excel spreadsheets and PostgreSQL databases. Python Flask supports API development for streamlined communication between the frontend and backend of the web app, while Node.js handles the app’s server-side operations. This data-driven approach intends to reduce hiring costs, improve intern placement accuracy, and support Fidelity’s goal of talent acquisition as well as operational efficiency
AI-Assisted Optimization of Pelton Turbine Designs for Low-Flow Desalination
No full textThis Major Qualifying Project presents the development of an interactive web application aimed at optimizing Pelton turbine parameters through the integration of Large Language Models (LLMs). The application, built with Python, Django, and OpenAI’s APIs for GPT-4 and DALL-E, enables users to input experimental conditions and receive optimized turbine design suggestions. Leveraging GPT-4, the application dynamically generates Python functions to calculate critical parameters, including bucket depth, bucket angle, and turbine diameter. Furthermore, the DALL-E API provides visual representations of the turbine designs based on the optimized parameters. Hosted on Heroku, this platform exemplifies the potential of AI-driven tools in engineering workflows, streamlining the design and optimization process for small-scale Pelton turbine systems and offering an accessible interface for iterative experimentation
Renovation of WPI Townhouses for Community Space
No full textWorcester Polytechnic Institute has invested time and resources into exploring potential redesigns and renovations of the Townhouses to increase the housing option’s desirability. This Major Qualifying Project focuses on designing and renovating the Townhouses to incorporate a community space using architectural design and structural engineering analysis. The scope of the proposed renovation includes the design of architectural programming and concepts, the structural schemes of a wood structure and a wood-steel hybrid structure, and mechanical systems based on changes to the building envelope and building thermal calculations
Teaching Functional Programming in High School with the Wipple Programming Language
No full textFunctional programming presents an alternate way of thinking computationally and offers a new challenge to students familiar with procedural programming. Based on my programming language Wipple, I designed and implemented a four-week course where technical high school students explored functional topics including iteration and pattern matching. The course’s three learning objectives emphasize self-directed learning through active reading of documentation, programming exercises, class discussion, and personal reflection. My daily observations and interactions throughout the course surfaced student misconceptions and breakthroughs related to planning and ordering, use of vocabulary, and modeling of control flow. These insights will inform future work on educational programming environments and complementary instructional materials, especially related to the presentation of error messages, with the goal of fostering a learner identity in all students
Software Tool to Process Data from Multiple LiDAR
No full textThis project developed a MATLAB-based software tool to identify the positions of elevator cables within a shaft using 2D LiDAR (Light Detection and Ranging) scan data. The primary goal was to create a non-invasive detection system that identifies consistent local minima in distance readings as indicators of suspended cables. Emphasis was placed on designing a robust masking and filtering pipeline that could suppress environmental noise, remove wall reflections, and preserve only meaningful geometric features relevant to cable detection. This preprocessing significantly enhanced detection reliability by aligning key features across multiple scan passes. Scan data was provided from a prior project and included five LiDAR configurations: the SICK R2000 (via both ROS and Windows), LSLiDAR N10 Plus, and SLAMTEC RP A1M8 and A2M8. All tests were performed under a fixed shaft setup labeled "3200_50", with the LiDAR positioned at a known distance, 3200 mm away from the far wall, and cables suspended in predetermined locations, 50 mm away from the far wall. Each LiDAR conducted three scans at the same position across the shaft, targeting the far wall. Cable positions were marked when multiple scan passes produced aligned local minima. Results showed that the SICK R2000 operating through ROS provided the clearest and most consistent detection. The system proved feasible for industrial use, especially when paired with high-precision LiDAR and ROS-integrated data acquisition
STEAM at Tower Bridge
No full textWe conducted research to determine the best practices for STEAM implementation for Tower Bridge’s exhibits. The primary method we used was interviews, which we conducted with staff at several museums to determine the current best practices in the museum and heritage site industry. We asked questions with the aim of understanding how they implement STEAM within their organization and how they select learning objectives. We also interviewed staff at Tower Bridge with similar questions, but with the addition of questions to help understand how they operate. As a secondary means of data collection, we also observed exhibits, both at Tower Bridge, and at other STEAM museums to compare and contrast methods of conveying information. Finally, we sat in on educational workshops at Tower Bridge, to further our understanding of how they currently apply STEAM. With this, we amassed a large collection of notes taken by hand, as well as audio recordings of interviews. With the audio interviews, we used a large language model to identify patterns between interviewees, in order to supplement our own takeaways. With this, we were able to categorize our qualitative data to create appropriate suggestions for Tower Bridge
Developing an E-Commerce Inventory System for the Ashland ReStore
No full textThe Ashland ReStore had implemented an e-commerce program to improve revenue, but lacked a comprehensive inventory system to support it. Drawing on over 608 hours of fieldwork, along with expert stakeholder interviews, we designed an updated, user-friendly inventory system to address this need. Employing an iterative design model, multiple rounds of feedback and user testing were conducted on the system. Upon completion of the project, the new system was implemented, and supporting documentation, including training manuals, was created and provided to the ReStore. By streamlining online operations, the system will expand its reach, increase revenue, and ultimately support Habitat for Humanity’s mission of making housing more affordable and accessible for those in need
To Tree or Not to Tree? Community Perspectives and Sustainable Solutions for Christmas Decorations
No full textThis project explores the feasibility, community interest, and sustainability of a winter holiday celebration at the Commonside Community Development Trust’s New Horizon Centre. The team conducted background research, semi-structured interviews, and surveys with staff, volunteers, patrons, trustees, and local stakeholders to analyze the communities’ opinions on their preferences for a winter celebration. Preliminary technical research included data on festive tree types, as well as lighting display options. Findings revealed that desired community gatherings fostered a sense of community pride. Based on these insights, this project recommends a culturally sensitive and environmentally sustainable light show surrounded by a community event