Seminar Teknologi dan Sains Nasional - STARNAS 2025

Audit Energi pada Bangunan Komersial Pusat Perbelanjaan Continent Jalan Koperasi Aikmel, Lombok Timur

Alimin Alimin — 2025-12-11

An energy audit was conducted on the commercial building “Continent,” located on Jalan Koperasi, Aikmel, with the objective of evaluating the level of electrical energy consumption and identifying potential energy efficiency improvements. The audit method included recording electrical loads, estimating equipment operational hours, calculating energy consumption in kWh, and analyzing the Energy Consumption Intensity (ECI) for each area. The audit results indicate that the monthly electrical energy consumption reaches 4,031.33 kWh with an estimated cost of Rp 5,824,068. The overall ECI value of 2.52 kWh/m²/month is categorized as efficient; however, the office area shows an ECI value of 13 kWh/m²/month, which exceeds the efficiency standards set by the Ministry of Energy and Mineral Resources. Based on these findings, several energy efficiency recommendations were formulated, including optimizing the lighting system, adjusting equipment operating hours, replacing conventional lamps with LEDs, installing motion sensors, and performing routine maintenance on electrical equipment. If all recommendations are implemented, it is estimated that energy savings of 1,686.17 kWh per month can be achieved, with a potential cost saving of Rp 2,436,009. Therefore, this audit concludes that there are significant opportunities for improving energy efficiency, particularly in areas with high energy consumption.

Integrasi Artificial Intelligence dalam Layanan Kefarmasian untuk Meningkatkan Kepatuhan Terapi pada Pasien Tunanetra dengan Tuberkulosis: Suatu Kajian Literatur

Fransiskus Samuel Renaldi — 2025-12-12

Tuberculosis remains a leading infectious disease globally, with medication non-adherence as a critical barrier to successful treatment. Visually impaired patients face additional challenges in medication management due to inability to read drug labels, thereby increasing the risk of medication errors and non-adherence. Artificial Intelligence (AI) presents innovative solutions through voice-based systems and adaptive technologies tailored to inclusive pharmaceutical care. This systematic literature review examines the efficacy, implementation opportunities, and challenges of AI-integrated pharmaceutical services in improving medication adherence among visually impaired tuberculosis patients. A comprehensive systematic review following PRISMA guidelines was conducted searching PubMed, Scopus, ScienceDirect, and Google Scholar (2018-2025). Twelve studies met inclusion criteria involving RCTs, protocols, and qualitative research examining AI applications in medication adherence and pharmaceutical care for vulnerable populations. AI technologies including voice assistants, natural language processing, smart pillboxes, and computer vision demonstrated 20-40% improvement in medication adherence. VDOT achieved 96.1% treatment completion rates compared to 94.6% for conventional DOT. However, accessibility barriers including limited smartphone ownership (30.4%) and demographic disparities in technology adoption remain significant. While AI-based interventions show promise for inclusive pharmaceutical services, successful implementation requires multidisciplinary collaboration, robust data security protocols, sustained healthcare provider training, and policy support ensuring equitable access for disabled populations.

Optimasi Model Random Forest pada Sistem Deteksi Serangan Jaringan IoT dengan Seleksi Fitur dan Hyperparameter Tuning

Fayruz Rahma — 2025-12-12

The Internet of Things (IoT) has become an integral part of various aspects of life, from smart homes to critical infrastructure. However, the increasing adoption of IoT also brings significant security challenges, including the growing number of cyberattacks. Therefore, intrusion detection in IoT networks is crucial for identifying and preventing potential threats. This study develops an attack detection model using the Random Forest (RF) algorithm, trained with the ToN_IoT dataset. To enhance model efficiency and accuracy, feature selection is applied using Recursive Feature Elimination (RFE), while parameter optimization is performed through Grid Search with 3-fold cross-validation. The resulting model exhibits high performance, with all evaluation scores exceeding 0.99 based on the Confusion Matrix and AUC-ROC, demonstrating its effectiveness in detecting attacks.

Transformasi Inovasi Perguruan Tinggi: Dampak Knowledge Sharing pada Inovasi Dosen dalam Mendukung Kemandirian dan Ketahanan Bangsa

Susan Febriantina — 2025-12-13

Innovation is essential for higher education institutions to adapt to rapid changes and enhance lecturers’ performance. One of the key factors that supports innovative work behavior (IWB) is knowledge sharing, which remains challenging in public universities with strong bureaucratic structures. This study aims to examine the influence of knowledge sharing on IWB among civil servant lecturers in higher education institutions under LLDikti Region III Jakarta. A quantitative causal-explanatory survey was conducted with 322 respondents selected through proportional random sampling. The findings indicate that knowledge sharing has a positive and significant effect on IWB, with a path coefficient of 0.370. The knowledge donating dimension emerged as the strongest contributor, reflecting lecturers’ high willingness to share insights through digital platforms. Meanwhile, knowledge collecting showed a lower score, indicating limited proactiveness in acquiring knowledge from colleagues. These results highlight that fostering a strong knowledge-sharing culture is crucial for stimulating lecturers’ innovative behaviors. In conclusion, institutions must strengthen collaborative ecosystems, enhance knowledge-sharing platforms, and build a work climate that actively encourages information exchange to support innovation in higher education.

Uji Sifat Fisik Sediaan Suspensi Amoxicillin dengan Menggunakan Pati Biji Durian (Durio zibethinus L.) Lokal sebagai Suspending Agent

Ade Irma Fitria Ningsih — 2025-12-13

Suspension is a liquid preparation that contains solid particles that do not dissolve in the liquid phase, so to dissolve the solid substance a suspending agent is required. Durian seeds are a type of waste that has not been widely utilized. After being studied, durian seeds contain 43.6 grams - 46.2 grams of carbohydrates per 100 grams of seeds which are converted into glucose. The aim of this study was to determine the effectiveness of local durian seed powder as a suspending agent at concentrations of 4%, 6%, and 8%. The research design used in this study was an experimental evaluation study. The results of the study showed that organoleptic tests at concentrations of 4%, 6% and 8% met the requirements for physical properties testing. The results of the pH test at concentrations of 4%, 6% and 8% on the first day were 5.0; 5.2; and 5.6, respectively, while on the sixth day the pH values were 5.2; 5.1 and 5.5, respectively. The results of the specific gravity test at concentrations of 4%, 6% and 8% showed results of 1.052 g/mL; 1.074 g/mL; and 1.096 g/mL, respectively. The results of the viscosity test at concentrations of 4%, 6% and 8% showed results of 51.3 cP; 54.3 cP; and 58.4 cP, respectively. The results of the sedimentation test at concentrations of 4%, 6% and 8% showed results of 0.11; 0.19; and 0.86. The results of the suspension redispersion test at concentrations of 4%, 6%, and 8% showed results of 83%, 86%, and 90%, respectively. From the results of the study, it can be concluded that the physical properties test of amoxicillin suspension preparations using local durian seed starch as a suspending agent only at a concentration of 8% met the physical properties test requirements.

Isolation of Escherichia Coli Flag Gene for Transformation

Muhammad Sidrotullah — 2025-12-14

The term "FLAG tag" refers to a hydrophilic peptide (DYKDDDDK) made up of eight residues that has an enterokinase cleavage site and is intended for the immunoaffinity purification and identification of recombinant proteins and strong specificity for monoclonal antibodies. Its hydrophilicity and tiny size help with efficient surface localization, which reduces disruption to protein activity. The method of this research is experimental. The experiment result will be described. Some processes will be done by polymerase chain reaction (PCR) to amplify the gene and then it will be extracted to pure the targeted gene. This research was successful to isolate Escherichia coli FLAG gene for transformation. This gene has important role to encode the flagella in the bacteria.

Penerapan Pembelajaran Mendalam pada Materi Menulis Cerpen Berbantuan Kecerdasan Buatan ChatGPT

Lalu Sirajul Hadi — 2025-12-17

Short story writing is an important literacy skill for junior high school students, yet many of them still struggle to develop coherent plots, shape convincing characters, and generate meaningful ideas. These challenges show the need for learning approaches that strengthen students’ reflective and analytical thinking. This study describes the implementation of a deep learning approach supported by ChatGPT in short story writing lessons and examines its influence on students’ writing performance and engagement. The research employed a descriptive qualitative method involving 40 purposively selected ninth-grade students at SMP Negeri 2 Mataram. Data were collected through classroom observations, questionnaires, and interviews with Indonesian language teachers. The results show clear improvement in students’ ability to organize story elements, develop characters, and refine diction after using ChatGPT. The tool also helped them explore ideas more confidently, which was reflected in an average score increase of 10–12 points. Teachers observed that the learning process became more reflective and meaningful, with students showing greater participation in shaping their narratives. Overall, the study concludes that integrating deep learning with AI-based assistance effectively enhances students’ short story writing skills and encourages more exploratory and reflective thinking.

Strategi Integratif Pemanfaatan AI Generatif dalam Pendidikan Tinggi untuk Meningkatkan Ketahanan Akademik dan Literasi Digital Nasional

Maria Atik Sunarti Ekowati — 2025-12-17

The integration of generative artificial intelligence (GenAI) in higher education presents a transformative opportunity to enhance academic resilience and digital literacy across Indonesia. This paper explores a multidisciplinary framework combining educational technology, cognitive science, and policy analysis to assess the impact of GenAI tools on learning personalization, academic productivity, and institutional preparedness. Using a mixed-methods approach, we analyze pilot implementations across three universities and propose a strategic roadmap for scalable adoption. Findings indicate that GenAI can significantly improve feedback cycles, adaptive learning, and research efficiency, provided ethical, infrastructural, and pedagogical challenges are addressed