Engineering Headway Vol. 27

Abstract: The increase in privacy concerns and the introduction of privacy and data protection legislation compel organisations to reevaluate their practices regarding traditional machine learning. The aggregation and management of users’ private data on the central server may contravene regulations if not properly administered. Federated learning provides a technique that eliminates the necessity of uploading users’ data to the server. It facilitates substantial learning by collaboratively training on each client’s devices and pooling the model gradient changes. Federated learning, augmented with a proxy as an intermediary and encrypted model parameters, will enhance anonymity, privacy, and data protection against malicious threats, including membership inference adversaries. Nonetheless, encrypted data incurs costs for customers’ communication and data size that exceed twice the original size. Our paper seeks to resolve these issues. We present two secure approaches for effective communication in an anonymous encrypted federated learning framework as our contribution. Additionally, our experiments demonstrated that it is feasible to attain equivalent communication costs as in non-encrypted scenarios. We provide recommendations in the conclusion for the effective implementation of privacy-preserving federated learning in the area of personal devices.

Abstract: Property intrusion remains a global concern, often driven by motives such as theft. While human-monitored CCTV systems are commonly employed for prevention, they are inherently limited by human capabilities and challenged by factors such as the increasing number of areas requiring surveillance and other factors. Advancements in Machine Learning (ML) offer promising solutions to overcome the limitations of human-monitored CCTV systems. Object detection, a potent ML method, enables real-time identification and tracking of objects through CCTV cameras. RT-DETR is an example of the excellent object detection model in terms of both accuracy and inference speed. The intention of this research is to implement RT-DETR on a web-based application to prevent property intrusion with such limited dataset quantity, for live video stream using NVIDIA RTX 3050 Mobile Laptop, which is a common and often considered as a mid-range GPU hardware. In addition to customizable surveillance zone definition, the application incorporates features to minimize false alarms, ensuring a more reliable and efficient security solution. Additionally, the application is equipped with automatic video recording as evidence of intrusion evidence.

Abstract: Cloud computing technology has rapidly advanced and has become a business necessity. Among its services is Container as a Service (CaaS), which offers advantages in application development due to its lightweight nature, ease of deployment, and scalability management. However, container management is complex and requires specialized skills. Traditional tools are often insufficient for simplifying container management and monitoring. Therefore, this paper proposes a CaaS architecture and its management platform to make container management easier for users. The proposed architecture utilizes Linux Containers (LXC), implemented through the Proxmox VE virtualization platform. For container management, a web-based platform is proposed, developed using NextJs. The web interface is designed to be simple, enabling even users without specialized skills to manage containers efficiently while still offering robust features for container management. The proposed CaaS architecture and management platform have been tested and function effectively, aiming to simplify the deployment and management of CaaS for users.

Abstract: A music chart is one way to measure the success and popularity of a song. One of the companies that presents music charts is Billboard Publication which serves as a critical reference point. However, many popular songs struggle to maintain longevity on the Billboard chart. This study focuses on predicting song longevity on the music charts, specifically the Billboard chart. The model incorporates characteristic data from previously charted popular songs on the Top 100 Billboard Chart and additional attributes from Spotify to ensure accurate predictions. The findings of this research will offer valuable insights to upcoming artists and producers by identifying the attributes they must focus on improving to enhance the popularity’s longevity of their music. Four machine learning models were utilized: Random Forest, Logistic Regression, Neural Network, and XGBoost. The tuned Random Forest model achieved an overall metric average of approximately 91.3%, followed by XGBoost with around 89.9%. These results demonstrate the effectiveness of decision tree models for this prediction task. Furthermore, artist-popularity, loudness, song-duration-ms, instrumentalness, and speechiness proved significant in this context.

Abstract: Distributed Software Development (DSD) or Global Software Development (GSD), today GSD is popular among practicing software developers. This is because GSD has advantages such as faster lead times, a wider range of superior workforce, and increased project cost efficiency. However, GSD also has challenges, including the implementation of RCM. Implementing RCM in GSD requires special attention because GSD has a working model that is applied at different locations, times and cultures for each project team. Poor implementation of RCM can result in project failure. The success of implementing RCM is influenced by the Critical Challenge Factor (CCF) and Critical Success Factor (CSF). In this paper, we conducted a Systematic Literature review to identify frequently occurring CCFs, looking at what methods are frequently used in identifying CCFs from previous research. In this paper, we conducted a Systematic Literature Review to identify frequently occurring CCFs, looking at what methods are frequently used in identifying CCFs from previous research. The results of the review found that as many as 29% of research used Systematic Literature Review (SLR), and in 52% of research, SLR was combined with several other approaches (Interview, Questionnaire Surveys, Expert Suggestion, or Observation). 14% of studies used an interview approach, and 5% used questionnaire surveys only. It means that SLR is the approach most often used to identify CCF, whether SLR is used alone (without combining it with other approaches) or in combination with other approaches. Furthermore, CCF, with the highest intensity of appearance in each literature, examines the fields of Communication and Coordination, namely at 24%, Social Culture at 12%, Management and Knowledge Sharing, and Time Zone difference at 6%. The approach used in identifying CCF is very dependent on the research objectives and research object. Therefore, choosing the right approach is expected to help practitioners to identify CCF better so that risk mitigation plans from CCF can be formulated appropriately.

Abstract: Cursive Chinese calligraphy in Taiwan is a traditional yet still prevalent art form. Due to the high degree of variation and the fact that most cursive script is derived from historical documents, the available data is relatively limited, posing a significant challenge for AI (artificial intelligence) models. In this study, we combine quantum computing with diffusion models to generate images of Chinese cursive characters. Diffusion models have recently been proven to perform better than other generative models when working with small datasets. Moreover, the integration of quantum computing reduces training costs and enhances generation performance. The results in this paper demonstrate the potential of quantum computing in conjunction with generative AI, which is applicable to interdisciplinary needs in design, artistic creation, and cultural preservation. In future work, we will delve deeper into noise-related issues and propose possible solutions.

Abstract: With the progress of digital transformation (DX) with the Internet, web application systems have become common at various situations in daily lives. Then, web-client programming using HTML, CSS, and JavaScript for dynamic web pages has increased the importance for students, particularly in information technology majors. However, in most university around the world, the corresponding courses are not offered due to limited time and teachers. As a result, the development of a self-learning tool has been strongly demanded. In web-client programming, the interactions between the three languages must be understood and mastered, to make programs of interactive and dynamic web pages. Then, as exercise problems in the self-learning tool, we have studied the behavior understanding problem (BUP). A question in a BUP instance requests to answer the key element in the given source code, which corresponds to the behavior described in the statement. The correctness of any answer is marked through string matching. To generate a proper BUP instance, the rules of selecting the key elements from the source code must be clarified. In this paper, we present the key element selection rules by analyzing interactions between HTML/CSS codes and JavaScript codes. Through observing manually generated BUP instances, we find that the rules should be made for the six categories: 1) JavaScript reserved word, 2) JavaScript library class/method, 3) JavaScript identifier, 4) Id name, 5) CSS syntax, and 6) text message. For the preliminary evaluation, we applied the proposed rules to ordinary web-client programming codes and confirmed their correctness. As the next step, we will implement the program of applying the rules automatically and verify the validity of our proposal.

Abstract: Nowadays, game programming is a dynamic and rapidly evolving domain, driven by its exceptional popularity and continuing growth in technology and entertainment sectors. As gaming industries expand across diverse platforms and audiences, the demand for proficient game developers has been risen significantly. In response, the need for accessible and effective learning resources is leading to a surge of online courses, tutorials, and learning tools. Previously, we have presented the element fill-in-blank problem (EFP) in the programming learning assistant system (PLAS) for self-study of basic web-client programming using HTML, CSS, and JavaScript by novice students. An EFP instance requests to fill in the blanks of the given source code by referring to the screenshots of the corresponding web page. The correctness of any answer is marked through string matching. Tags, function names, or text messages may be blanked in the code. In this paper, we present EFP for self-study of web-client game programming. By solving proposed EFP, it is expected that students not only gain hands-on experience in game programming but also learn in-depth unique programming styles and grammar concepts there. For preliminary evaluations, we generated three EFP instances using source codes for web-client games and assigned them to 20 students in State Polytechnic of Malang, Indonesia. Their solution results confirmed the validity of the proposal.

Abstract: Industrial monitoring and control systems, using PLC and SCADA, provide us with real-time information about the performance of each machine and process. This monitoring system allows us to continuously monitor and control its operations. The application of SCADA systems in the automotive manufacturing industry is the first step towards the Industrial Revolution 4.0 based on machine learning and artificial intelligence. There is a problem with recording production results in one area of the automotive industry. When carrying out the recording process, the manual processes are still used and production timeline stop problems are not well controlled, resulting in inefficient time consumption. Next, research and improvement were carried out by creating a production monitoring system (PMS) that could be used to monitor production results in that area. This research was developed with the integration of PLC (Programmable Logic Controller) control, SCADA software, and database software. In this case, we designed the Mitsubishi PLC program, created the programming, displayed it with the CIMON SCADA application, and created the database system. This system has been proven to make it easier for the production team to monitor production quantities, monitor production line stops, and increase production efficiency.