Implementasi Prinsip Explainable Artificial Intelligence dalam Desain Antarmuka Sistem Monitoring Bobot Ayam untuk Meningkatkan User Trust dan Understandability

Maritza Angelina Az Zahra, Divi Galih Prasetyo Putri, S.Kom., M.Kom., Ph.D.

2026 | Tugas Akhir | D4 Teknologi Perangkat Lunak

Pemanfaatan Artificial Intelligence (AI) pada aplikasi monitoring bobot ayam semakin berkembang dalam mendukung analisis data dan pengambilan keputusan di sektor peternakan, namun sistem yang masih bersifat black box menyebabkan rendahnya pemahaman dan kepercayaan pengguna. Sejumlah penelitian sebelumnya telah mengkaji pengalaman pengguna pada sistem berbasis AI, tetapi integrasi prinsip Explainable Artificial Intelligence (XAI) ke dalam desain antarmuka untuk membantu pengguna memahami cara kerja sistem AI masih belum banyak dilakukan. Oleh karena itu, penelitian ini bertujuan untuk menganalisis pengaruh penerapan XAI terhadap user trust, understandability, dan usability. Penelitian dilakukan pada aplikasi monitoring bobot ayam dengan membandingkan antarmuka eksisting dan versi modifikasi berbasis XAI menggunakan pendekatan within-subject dengan melibatkan 16 partisipan. Pengukuran dilakukan menggunakan Trust in Automation Scale (TiAS) untuk user trust, instrumen understandability yang dikembangkan berdasarkan sintesis beberapa literatur, serta System Usability Scale (SUS) untuk usability. Analisis data dilakukan melalui uji normalitas, uji hipotesis dengan paired-samples t-test, serta perhitungan effect size dengan Cohen’s d untuk menilai kekuatan pengaruh. Hasil analisis menunjukkan adanya peningkatan signifikan pada seluruh variabel dengan nilai signifikansi < 0>

The use of Artificial Intelligence (AI) in chicken weight monitoring applications has continued to grow in supporting data analysis and decision-making processes in the livestock sector. However, the black-box nature of many AI systems often leads to low levels of user understanding and trust. Previous studies have examined user experience in AI-based systems, but the integration of Explainable Artificial Intelligence (XAI) principles into interface design to help users understand how AI systems work remains limited. Therefore, this study aims to analyze the effect of implementing XAI on user trust, understandability, and usability. The study was conducted on a chicken weight monitoring application by comparing the existing interface with a modified XAI-based interface using a within-subject approach involving 16 participants. Measurements were conducted using the Trust in Automation Scale (TiAS) for user trust, an understandability instrument developed based on the synthesis of several studies, and the System Usability Scale (SUS) for usability. Data analysis was performed through normality testing, hypothesis testing using paired-samples t-tests, and effect size calculation using Cohen’s d to measure the magnitude of the effect. The results showed significant improvements across all variables, with significance values of p < 0>

Kata Kunci : chicken weight monitoring, Explainable Artificial Intelligence (XAI), user trust, understandability, usability