Universitas Atma Jaya Yogyakarta (UAJY): Open Journal Systems
Not a member yet
9386 research outputs found
Sort by
Pengering Kerupuk dengan Mesin Kompresi Uap untuk Mengurangi Pencemaran Lingkungan
Full text linkPabrik kerupuk merupakan tempat produksi makanan kecil kerupuk yang menjalankanserangkaian proses produksi yang cukup panjang. Setelah tersedia bahan pokok yang diperlukan, maka dimulai membuat adonan bahan kerupuk yang siap cetak, mencetak, mengkukus, mengeringkan, dan diakhiri dengan menggoreng. Salah satu bagian proses pembuatan kerupuk yang membutuhkan perhatian khusus adalah kegiatan pengeringan. Hal ini dikarenakan akan menentukan kualitas dan kuantitas kerupuk yang bisa diproduksi setiap hari. Proses pengeringan yang dilakukan selama ini adalah dijemur di bawah Terik sinar matahari dan dioven dengan bahan bakar kayu. Kendala yang dihadapi dalam proses pengeringan kerupuk dengan dijemur adalah cuaca yang tidak menentu dan hanya dapat disiang hari. Kendala yang dialami jika dikeringkan menggunakan oven kayu bakar adalah suhu yang tinggi dan pencemaran udara ke lingkungan pabrik. Untuk mengatasi adanya gangguan lingkungan tersebut, maka kegiatan pengabdian kepada masyarakat ini dilakukan pembangunan sistem pengering kerupuk metode kompresi kerupuk dengan sumber energi listrik. Luas ruang pengering 2,5x4 m2 dengan kapasitas mesin pengering 1 PK mampu mengeringkan kerupuk selama 11 jam
Model Prediksi Cacat Produk sebagai Early Detection System pada Quality Control Manufaktur
Full text linkDalam sistem manufaktur modern, produk cacat merupakan salah satu bentuk pemborosan yang berdampak langsung pada efisiensi produksi, biaya, dan kualitas produk. Pada sistem produksi massal, keterlambatan dalam mendeteksi cacat berpotensi menghasilkan produk tidak sesuai spesifikasi dalam jumlah besar. Namun, pemanfaatan model prediksi berbasis data produksi untuk mendeteksi potensi cacat secara dini masih relatif terbatas dan belum banyak dikaji secara komparatif. Penelitian ini bertujuan untuk mengembangkan model awal prediksi tingkat cacat produk manufaktur menggunakan pendekatan machine learning. Metode yang digunakan adalah Random Forest (RF) dan Multi-Layer Perceptron (MLP) dengan memanfaatkan dataset manufaktur terbuka, serta dilakukan analisis feature engineering dan hyperparameter tuning. Kinerja model dievaluasi menggunakan train accuracy dan test accuracy serta selisihnya sebagai indikator overfitting. Hasil menunjukkan bahwa RF menghasilkan akurasi prediksi yang lebih tinggi, terutama setelah hyperparameter tuning, dengan penurunan overfitting yang signifikan. Sementara itu, MLP menunjukkan stabilitas yang lebih baik tetapi dengan akurasi yang lebih rendah. Penelitian ini menunjukkan bahwa model prediksi berbasis data produksi berpotensi menjadi alat bantu awal bagi Quality Control dalam mendeteksi risiko cacat secara dini di lingkungan manufaktur
PERANCANGAN USER INTERFACE FITUR NEW USER JOURNEY DAN REFERRAL PROGRAM APLIKASI SUPERAPP MENGGUNAKAN METODE DESIGN THINKING
Full text linkAplikasi SuperApp merupakan sarana utama perusahaan PT. Krakatau Karya Abadi dalam menyediakan informasi dan layanan kepada pengguna. Perusahaan menyajikan informasi, fitur, produk dan layanan, dan berbagai hal lainnya kepada pengguna melalui aplikasi SuperApp. Diketahui aplikasi belum mencapai target monthly active users selama quarter tiga, sehingga perusahaan berencana menambahkan fitur baru pada aplikasi yaitu new user journey dan referral program. Pengembangan dua fitur ini memerlukan tampilan user interface yang sesuai dengan kebutuhan dan harapan pengguna. Namun, tampilan user interface untuk dua fitur ini belum ada, sehingga tujuan dari penelitian ini adalah untuk merancang user interface fitur new user journey dan referral program yang menarik dan user-friendly. Metode yang digunakan yaitu design thinking dengan pendekatan pada manusia berdasarkan karakteristik, psikologis, dan persepsi manusia. Hasil dari penelitian ini adalah prototype yang telah diuji dan dievaluasi menggunakan Single Ease Question dengan rata – rata nilai lebih dari 6 yang menunjukkan bahwa desain dapat dikategorikan mudah digunakan dan pengguna merasa puas dengan desain prototype yang telah dibuat
Sistem Rekomendasi Kafe Di Kota Palangka Raya Menggunakan Metode Collaborative Filtering: Indonesia
No full textPertumbuhan jumlah kafe di Palangka Raya memberikan dampak dimana konsumen seringkali kebingungan dalam memilih kafe yang sesuai dengan preferensi pribadi konsumen. Penelitian ini bertujuan membangun sistem rekomendasi berbasis web yang mempersonalisasi rekomendasi kafe menggunakan pendekatan User-Based Collaborative Filtering (UBCF).
Preferensi pengguna dimodelkan dalam matriks penggun dan kafe berdasarkan total harga menu yang disukai, lalu diproses dengan mean-centering untuk perbandingan yang lebih adil. Kemiripan antar pengguna dihitung menggunakan cosine similarity dan tetangga terdekat ditentukan dengan menggunakan K-Nearest Neighbors (KNN). Skor kemiripan ini kemudian digabungkan dengan sinyal tambahan seperti riwayat kunjungan, menu favorit, rating, dan sentimen menggunakan teknik weighted sum untuk menghasilkan 6 rekomendasi terbaik.
Evaluasi sistem dilakukan dengan dua aspek, yaitu evaluasi akurasi prediksi dan evaluasi kualitas urutan (ranking) rekomendasi yang diberikan. Pengujian akurasi menggunakan skema 5-fold cross validation menghasilkan rata-rata RMSE 0.2063 dan MAE 0.1748, melampaui nilai baseline penelitian terdahulu. Sementara itu, evaluasi kualitas ranking menunjukkan hasil positif dengan Precision@1 sebesar 0.8093, Recall@10 sebesar 0.8995, dan nDCG@1 sebesar 0.8093. Nilai F1-score tertinggi tercatat pada k=3, menandakan keseimbangan optimal antara presisi dan recall pada tiga rekomendasi teratas
Entrepreneurship and Marketing Communication Mentoring for Post-Migrant Families in Resapombo Village, Blitar, East Java
No full textThis mentoring program is addressed to former migrant families in Resapombo Village, Blitar, East Java, in response to the economic issue. The entrepreneurship and marketing communication program aims to create a resilient economy in remote areas. The methods used in this mentoring are socialization, workshops/training, cadre recruitment, and entrepreneurial practice. The primary materials presented in this mentoring program include building entrepreneurial thinking and spirit, canvas business models, creating brands and marketing communication planning, saving culture, and family financial management. The recruitment of cadres is essential in this activity to maintain the program's sustainability. Participants were asked to put their business ideas into practice, and then they were given insights into obtaining business capital for the last topic. It is designed to ensure that ideas can be realized and that economically resilient families can be realized
Pelatihan Literasi Media Digital dan Public Speaking bagi Komunitas Piwulang Becix Yogyakarta
No full textHomeschooling in Indonesia is recognized as an alternative form of education that provides flexibility in the learning process. However, many people still lack a clear understanding of its benefits and challenges, particularly in developing children's practical skills. This community service program aims to develop public speaking and digital media literacy skills among children and adolescents through activities implemented by the Piwulang Becix Community in Yogyakarta. The method used is an interactive approach, including audiovisual media training, social media literacy, and public speaking techniques. Children are introduced to various audiovisual technologies, such as green screens and virtual reality, to stimulate their creativity and engagement in learning. The results of this activity indicate an improvement in children's skills in using digital media more critically and responsibly, as well as increased confidence in speaking in public. However, several challenges are still found, especially anxiety when speaking in front of an audience among some participants. Overall, this program provides a more personal, contextual, and applicable learning experience related to digital media and public speaking. It is expected that this training can serve as adequate initial preparation for children to face the challenges of digital media more wisely and effectively in the future
Pemanfaatan Limbah Plastik menjadi Eco-Conblock sebagai Upaya Peningkatan Kapasitas Lokal dan Pemberdayaan Masyarakat Desa Ponggang
No full textAs a tourism village, Ponggang faces problems related to waste management that are currently not being handled properly. If not followed up, this will pose a threat to both public health and the potential for damage to the village's natural and environmental attractions. Therefore, this community service aims to empower community groups through the utilization of plastic waste into products with higher economic value (eco-conblock). The methods used in this service include outreach activities aimed at increasing awareness and understanding of partners about the concept of village tourism and training/mentoring activities that involve the active role of community members aimed at increasing technical skills and waste management governance. From the evaluation results, it can be seen that the aspect of community awareness and understanding of good waste management has significantly increased from 3.85 points (good) to 4.47 points (very good). Meanwhile, the aspect of skills and governance has increased significantly on average from 3.39 points (moderate) to 3.68 points (good). In addition, the community service program has also succeeded in building an integrated waste management facility that can later be used sustainably towards Ponggang becoming an independent tourism village
Adaptive Integration of Distributed Deep Q-Networks for Enhancing OLSR Routing in Dynamic Mobile Ad-Hoc Networks
Full text linkAdaptive routing in Mobile Ad-Hoc Networks (MANETs) poses considerable difficulty owing to the network's dynamic characteristics, lack of stable infrastructure, and swift topology alterations. The Optimized Link State Routing (OLSR) protocol provides a proactive routing mechanism via topology dissemination and MultiPoint Relay (MPR) selection. Nevertheless, it exhibits diminished responsiveness to real-time topology alterations, as it depends on periodic updates and does not explicitly account for link quality. This paper suggests the incorporation of the Deep Q-Network (DQN) methodology into OLSR as a reinforcement learning strategy to improve routing adaptability and efficiency. The DQN model employs network metrics like latency, ETX, buffer occupancy, and neighbor count as state inputs, with actions determined by Q-values obtained via environmental interactions. Simulations conducted with NS-3 and PyTorch demonstrate that OLSR-DQN enhances Packet Delivery Ratio (PDR) by as much as 20%, decreases delay by 15–25%, and markedly boosts throughput in dynamic MANET situations.
Keywords: MANET, OLSR, Deep Q-Network, adaptive routing, reinforcement learnin
Penentu Minat Penggunaan QRIS Di Tempat Ibadah: Manfaat, Kemudahan, Risiko, dan Kepercayaan
Full text linkThis study aims to analyze the determinants of interest in using QRIS in places of worship, namely perceived benefits, ease of use, risk, and trust. The phenomenon shows that students are interested in using QRIS for donations in places of worship. This study adopts a quantitative approach using primary data obtained through questionnaires distributed to students. The sampling method used non-probability sampling, while data analysis was performed using multiple linear regression with SPSS version 27. The results show that perceived benefits and ease of use do not significantly affect interest in using QRIS in places of worship. Conversely, perceived risk has a negative effect on interest in using QRIS in places of worship, meaning that the higher the perceived risk, the lower the interest in its use. Meanwhile, trust showed a significant positive effect on the interest in using QRIS in places of worship. These research results indicate that trust and risk perception are the main determinants of interest in using QRIS in places of worship, while perceived benefits and ease of use are not significant determinants
Penguatan Branding Produk Pertanian Unggulan Melalui Integrasi Teknologi Solar tunnel dryer di BUMDes Tamanmartani
Full text linkThe community service program at BUMDes Tamanmartani, Sleman Regency aims to strengthen the branding of leading agricultural products through the integration of Solar tunnel dryer technology. This collaboration is designed to enhance the capacity of BUMDes in managing local agricultural products, providing added value and improving their competitiveness in the market. Branding reinforcement is carried out by exploring the potential of village flagship products and developing marketing strategies based on innovative drying technology. The social entrepreneurship approach offers opportunities for villagers to collaborate, share knowledge, and optimize agricultural yields with improved quality. Problem-solving serves as the main method in addressing the challenges faced by BUMDes, including production management, marketing, and resource utilization. This is achieved through participatory discussions between the community service team, BUMDes administrators, and members of BUMDes. Furthermore, the program includes marketing training and the sustainable use of Solar tunnel dryer technology. Activities also cover the development of branding programs for agricultural products, the enhancement of BUMDes marketing skills, and the strengthening of distribution networks, enabling Tamanmartani’s agricultural products to compete at both local and regional levels