penikmatrumput/nasa-cmapss-rul

Modified CMAPSS Dataset (Turbofan Engine Degradation)

📘 Description

This dataset is a modified version of the NASA C-MAPSS (Commercial Modular Aero-Propulsion System Simulation) turbofan engine degradation simulation dataset. The modification was created by our team as part of a submission for RISTEK UI Datathon 2025, in conjunction with the predictive modeling work we developed.

Each entry in this dataset corresponds to one engine's operating cycle. Engines begin with varying degrees of wear and manufacturing variation, operate normally, and eventually experience degrading performance leading to system failure. The task is to predict Remaining Useful Life (RUL) based on sensor and operating condition readings.

📂 Dataset Subsets

The dataset is divided into four subsets, consistent with the original C-MAPSS structure:

Subset	Train Units	Test Units	Operating Conditions	Fault Modes
FD001	100	100	1 (Sea Level)	1 (High Pressure Compressor)
FD002	260	259	6	1 (High Pressure Compressor)
FD003	100	100	1 (Sea Level)	2 (HPC + Fan Degradation)
FD004	248	249	6	2 (HPC + Fan Degradation)

🛠️ Modifications

The original dataset has been preprocessed and extended for easier machine learning use:

• Added RUL column: Remaining Useful Life (RUL) has been precomputed for each cycle in the training set.
• Added headers: All columns are clearly labeled for usability.
• Converted format: Original space-delimited files are now provided as .csv files.

🧾 Data Fields

Each row corresponds to a single time step (cycle) for one engine unit. It includes 27 columns: unit_number, time_in_cycles, op_setting_1, op_setting_2, op_setting_3, sensor_1, sensor_2, ..., sensor_21, RUL

📈 Intended Use

• Benchmarking and training models for Remaining Useful Life (RUL) prediction
• Educational and research use in predictive maintenance and health monitoring

⚠️ Limitations and Considerations

• This dataset is simulated and may not fully reflect real-world engine behavior.
• Only includes two types of fault modes.
• Best used as a benchmark for research rather than direct deployment.

📚 Citation

If you use this dataset, please cite the original NASA dataset as well as our submission if relevant:

@inproceedings{saxena2008damage, title={Damage Propagation Modeling for Aircraft Engine Run-to-Failure Simulation}, author={Saxena, A. and Goebel, K. and Simon, D. and Eklund, N.}, booktitle={Proceedings of the 1st International Conference on Prognostics and Health Management (PHM08)}, year={2008}, address={Denver, CO} }

Dataset CMAPSS yang Dimodifikasi (Degradasi Mesin Turbofan)

📘 Deskripsi

Dataset ini adalah versi modifikasi dari dataset simulasi degradasi mesin jet turbofan NASA C-MAPSS. Modifikasi ini dibuat oleh tim kami sebagai bagian dari submisi untuk RISTEK UI Datathon 2025, seiring dengan model prediktif yang kami kembangkan.

Setiap entri merepresentasikan siklus operasional dari satu mesin. Mesin dimulai dalam kondisi normal (dengan variasi keausan awal), kemudian mengalami degradasi hingga akhirnya gagal. Tugas utama dari dataset ini adalah memprediksi Remaining Useful Life (RUL) berdasarkan pembacaan sensor dan kondisi operasi.

📂 Subset Dataset

Dataset dibagi ke dalam empat subset seperti pada struktur CMAPSS asli:

Subset	Unit Latih	Unit Uji	Kondisi Operasi	Mode Kegagalan
FD001	100	100	1 (Permukaan Laut)	1 (High Pressure Compressor)
FD002	260	259	6	1 (High Pressure Compressor)
FD003	100	100	1 (Permukaan Laut)	2 (HPC + Degradasi Fan)
FD004	248	249	6	2 (HPC + Degradasi Fan)

🛠️ Modifikasi

Dataset asli dimodifikasi untuk memudahkan pemrosesan dan pemodelan:

• Penambahan kolom RUL: Sisa siklus hidup (RUL) dihitung dan ditambahkan.
• Header kolom ditambahkan: Semua kolom diberi nama deskriptif.
• Format diubah: File teks dipisahkan spasi diubah menjadi .csv.

🧾 Struktur Data

Setiap baris merepresentasikan satu siklus operasional untuk satu unit mesin dengan total 27 kolom: unit_number, time_in_cycles, op_setting_1, op_setting_2, op_setting_3, sensor_1, sensor_2, ..., sensor_21, RUL

📈 Penggunaan yang Disarankan

• Melatih dan menguji model prediksi Remaining Useful Life (RUL)
• Penelitian dan edukasi dalam konteks predictive maintenance

⚠️ Keterbatasan

• Ini adalah data hasil simulasi, bukan dari mesin nyata
• Hanya mencakup dua jenis mode kegagalan
• Sebaiknya digunakan sebagai benchmark eksperimen, bukan untuk deployment langsung

📚 Sitasi

Jika Anda menggunakan dataset ini, mohon sitasi dataset asli NASA serta submisi kami jika sesuai:

@inproceedings{saxena2008damage, title={Damage Propagation Modeling for Aircraft Engine Run-to-Failure Simulation}, author={Saxena, A. and Goebel, K. and Simon, D. and Eklund, N.}, booktitle={Proceedings of the 1st International Conference on Prognostics and Health Management (PHM08)}, year={2008}, address={Denver, CO} }