No abstract

No abstract

Predicting slope failure is one of the most sought after feature from Slope Stability Radar (SSR). An accurate slope failure prediction will potentially give an ample time to manage risk related with slope stability, wherein the evacuation ofequipment or personal would be executed on a timely manner. The renownedmethod to predict failure among geo-mechanical practitioner is utilizing inversevelocity method, in which collapse will be predicted to happen when the extension of inverse velocity line is intercepted at predefined value that is usually only fractal above zero. The tenet of this method is, if one has acquired the knowledge of inverse velocity value from previous collapses, the next collapse could be predicted based on it with the pretext that both share the same nature and geological feature. The same can be said for predicting collapse based on velocity value. Set of maximum velocity value from several previous collapses will be averaged to determine predefined assumption to predict the next collapse. This paper will demonstrate an alternative method to predict collapse that will use velocity value instead of inverse velocity. This method is called SLO method as proposed by Azania Mufundirwa.This paper will specifically exemplify the practical steps to produce the failureprediction from slope stability radar data, and discuss the characteristic of theprediction yield by this method. Velocity chart with velocity calculation period of60 minutes is first established from particular pixel deemed as the one that showing the most distinguished progressive deformation trend. The velocity data will then be an exported and reprocess as such that the time data will be converted into unit time stamp number. The designated time stamp will then be accumulated, in which the onset of failure, will be regarded as time 0 reference. Log linear chart will be generated in which X-axis will be occupied by velocity value, while Y-axis will depict Velocity x Accumulated time (SLO chart). Collapse can subsequently be predicted by intercepting the predefined assumption of velocity during collapse with the log linear curve from the SLO chart. Two methods, mathematical & graphical, will be presented in this paper in order to give in depth understanding as to how one can predict collapse event with velocity value. Taking account on the study case from iron ore mining, SLO method yielded prediction of failuretime on 10:58 PM 31st January 2016, meanwhile the real failure occur on 11:32 PM 31st January 2016.

Provided with accurate and quasi real time deformation data, there are at least 2 methods that can be utilized to predict a slope failure. Inverse velocity method, coined by Fukuzono, aims at the interception of inverse velocity line to zero value at X time axis as the prediction of slope failure. More recent method called SLO, develop by Mufundirwa, puts emphasize on interception of acceleration regression line with X velocity axis. This paper is intended first and foremost to establish well-structured comparison between the two aforementioned methods. By using the same set of displacement data that show progressive deformation trend from Slope Stability radar, both SLO & Inverse Velocity method will be put into trial. Not only the accuracy of the failure prediction time, but also the comparison between the R2 attribute will be examine to reveal which method that yield better data statistically. One of the selected study case, from several which is presented on the paper, reveal that SLO method give failure prediction closer with the actual failure compared to Inverse Velocity method. The actual failure is happening at 21:59 AM January 1st 2016. SLO method generates failure prediction 10 minutes prior the actual failure, while Inverse Velocity generates failure prediction plus 68 minutes after the failure. R2 value for SLO method and Inverse Velocity method respectively are 0.710 & 0.630. Apart from this results comparison, a more in depth examination toward the nature of both methods delivers pro & con of each method. SLO method seems more accurate but having a constraint in which if there are no previous database of maximum velocity during collapse, prediction is almost impossible to make. Inverse Velocity method could address this flaw by projecting the inverse velocity line to zero value for the very least. Further explanation about the flaw and advantages of both methods will be conveyed in more detail on the later part of this paper.   Key words: Failure Prediction, SLO, Inverse Velocity, SSR  ABSTRAK Dengan adanya pengambilan data deformasi yang akurat dan mendekati “real time”, terdapat setidaknya dua metode yang dapat digunakan untuk memprediksi longsor. Metode inverse velocity, yang dikembangkan oleh Fukuzono, adalah metode yang menggunakan perpotongan grafik inverse velocity dengan titik nol sebagai acuan atau nilai dari prediksi longsor. Metode lain yang lebih baru dibandingkan metode inverse velocity adalah metode SLO yang dikembangkan oleh Mufundirwa. Metode ini lebih ditekankan pada perpotongan antara grafik akselerasi dengan nilai kecepatan pada sumbu X. Tujuan utama dari paper ini adalah penyajian perbandingan yang terstruktur antara kedua metode tersebut. Penelitian terhadap metode SLO dan inverse velocity menggunakan data deformasi progresif yang sama dari Slope Stability Radar. Tidak hanya keakuratan prediksi waktu longsor, tetapi perbandingan nilai R2 pun akan menentukan metode yang lebih efektif secara statistik. Pada salah satu studi kasus, dari beberapa kasus yang dibahas di paper ini, menunjukkan bahwa metode SLO memberikan prediksi waktu longsor yang lebih mendekati waktu longsor yang sebenarnya jika dibandingkan dengan metode inverse velocity. Longsor yang sebenarnya terjadi pada tanggal 1 Januari 2016, pukul 21:59. Metode SLO menghasilkan prediksi longsor 10 menit lebih awal dari waktu longsor yang sebenarnya, dimana metode inverse menghasilkan prediksi longsor 68 menit setelah waktu longsor. Nilai R2 untuk metode SLO dan inverse velocity adalah 0.71 dan 0.63. Di samping perbandingan kedua hasil di atas, pemahaman lebih mendalam tentang sumber dari kedua metode tersebut memunculkan hasil plus dan minus dari masing-masing metode. Metode SLO memang terlihat lebih akurat namun metode ini membutuhkan data kecepatan maksimal saat kejadian longsor sebelumnya. Jika tidak ada, maka prediksi hampir tidak mungkin untuk dibuat. Sebaliknya, kelemahan tersebut tidak terdapat pada metode inverse velocity karena dapat diproyeksikan pada titik nol. Penjelasan lebih dalam mengenai kelebihan dan kekurangan dari kedua metode tersebut akan dibahas selanjutnya pada paper ini. Kata kunci: Prediksi longsor, SLO, Inverse velocity, SSR

No abstract