Fenomena Microsleep pada operator Dump Truck di industri pertambangan merupakan salah satu tantangan keselamatan kerja krusial yang dipicu oleh karakteristik pekerjaan monoton, durasi kerja panjang, dan sistem kerja shift. Pengawasan manual dan pendekatan administratif konvensional memiliki keterbatasan dalam mendeteksi penurunan kewaspadaan secara Real-Time. Penelitian ini bertujuan untuk menganalisis angka kejadian Microsleep sebelum dan setelah penerapan sistem Defender berbasis AI serta mengevaluasi tingkat efektivitas teknologi tersebut dalam menurunkan angka kejadian Microsleep operator Dump Truck di PT X. Metodologi yang digunakan adalah metode campuran sekuensial eksplanatori (Mixed Methods Sequential Explanatory Design). Data kuantitatif diperoleh dari kuesioner berskala Likert yang disebarkan kepada 201 operator Dump Truck (HD) dan dianalisis menggunakan program SPSS, serta dilengkapi data sekunder laporan Fatigue Incident Record periode 2020–2025. Data kualitatif dieksplorasi melalui wawancara mendalam bersama 8 pengawas lapangan dan 5 personel Safety, Health, and Environment (SHE). Hasil penelitian menunjukkan bahwa sebelum implementasi sistem Defender (periode 2020–2023), angka kejadian Microsleep tergolong sangat tinggi dengan rata-rata 1.262 kejadian per tahun. Setelah penerapan sistem Defender berbasis AI, jumlah kejadian Microsleep turun drastis menjadi 415 kejadian pada tahun 2024 dan mencapai titik terendah sebesar 207 kejadian pada tahun 2025 (rata-rata 311 kejadian). Hasil analisis data sekunder membuktikan tingkat efektivitas sistem sebesar 75,36% dalam menekan kejadian Microsleep. Diskusi menekankan bahwa alarm visual dan audio Real-Time dari Defender secara signifikan mempercepat deteksi dini dan memicu respons pengawasan proaktif, meskipun tantangan teknis seperti False Alarm akibat debu atau APD masih ditemukan. Dengan demikian, penelitian ini menyimpulkan bahwa penerapan teknologi monitoring fatigue Defender berbasis AI terbukti sangat efektif menurunkan angka kejadian Microsleep operator Dump Truck di PT X dan memperkuat implementasi Fatigue Risk Management System (FRMS).

The phenomenon of Microsleep among Dump Truck operators in the mining sector presents a critical occupational safety challenge, primarily driven by monotonous tasks, extended working hours, and shift work schedules. Conventional manual supervision and administrative controls have proved inadequate for detecting alertness degradation in Real-Time. This study aims to analyze the Microsleep incidence rates before and after the implementation of the AI-based Defender system and to evaluate the effectiveness of this technology in reducing Microsleep among Dump Truck operators at PT X. The methodology employed a Mixed Methods Sequential Explanatory Design. Quantitative data were gathered through Likert-scale questionnaires distributed to 201 Dump Truck (HD) operators, analyzed via SPSS software, and supplemented by secondary corporate Fatigue Incident Records from 2020 to 2025. Qualitative insights were extracted from in-depth interviews involving 8 field supervisors and 5 Safety, Health, and Environment (SHE) personnel. The findings indicate that prior to the system's integration (2020–2023), the frequency of Microsleep was notably high, averaging 1,262 incidents per year. Following the deployment of the AI Defender system, Microsleep occurrences fell sharply to 415 in 2024 and reached an all-time low of 207 incidents in 2025 (averaging 311 events). Secondary data analytics revealed a 75.36% effectiveness rate in mitigating Microsleep occurrences. The discussion highlights that Defender’s Real-Time audio-visual alerts substantially expedited early detection and triggered proactive supervisory interventions, despite enduring technical constraints like dust-induced or PPE-induced False Alarms. In conclusion, this study demonstrates that the implementation of the AI-powered Defender fatigue monitoring system is highly effective, reducing Microsleep incidence among Dump Truck operators at PT X by 75.36% and bolstering the corporate Fatigue Risk Management System (FRMS).