Aircraft Flight Movement Anomaly Detection Using Automatic Dependent-Surveillance Broadcast (ADS-B) Data
The topic I want to discuss this time is about the utilization of Automatic Dependent Surveillance-Broadcast (ADS-B) data to determine (predict) the percentage of flight anomalies from its data analysis. The last flight history data of an aircraft plays an important role in identifying abnormal flight (anomalies) that occur. Aircraft flight movement anomaly detection is a technique to identify rare events or observations of changes in information that occur significantly.
ADS-B is a device or system of technology that monitors aircraft that receives sensor information from each aircraft. There are many community projects that utilize ADS-B, such as websites that offer live radar services using commercially available hardware (Strohmeier et al., 2014). The ADS-B technology uses a radio signal transmitted by the aircraft transponder (Mode-S). Position, speed, altitude, and aircraft identification are important information contained in each signal, where the signal (ADS-B Out) is broadcast to unknown receivers (ADS-B In) and has an average transmission rate of 0.5 seconds at 1090 MHz frequency (Gagliardi et al., 2017). In concept, if all aircrafts use ADS-B, then aircraft identification, monitoring, and the possibility of conflict avoidance (accidents) can be facilitated (Guterres et al., 2017).
The issues in this topic occur when accidents still happen in Indonesia, especially since this thesis is made coinciding with the fall of Sriwijaya Air 182 with callsign SJY182. Yet, if we look back, regulations regarding the use of ADS-B as a mandatory device for air transportation have already been implemented. Until now, the use of ADS-B has not been felt and technology needs to be improved to prevent and reduce aviation accidents that occur in Indonesia.
This study is an improvement of previous research conducted by Nanduri & Sherry, utilizing the Deep Learning method and ADS-B sensor data of aircraft to detect anomalies in the movement of aircraft from the X-Plane simulation game. We apply real-life aircraft accident incident data from the Flightradar24 community-owned ADS-B using Artificial Intelligence (AI) that focuses on Deep Learning (DL) models.
In the labeling process, the dataset first needs to be preprocessed, then analyzed based on the aircraft accident investigation report issued by aviation authorities. After that, the data is processed with sliding windows to standardize its characteristics resulting in data that is ready to be modeled.
The results of this study are promising to be applied to the aviation industry, because the ADS-B device can be used as a backup radar in
monitoring and detecting aircraft movement anomalies. In addition, for future research, the model can be implemented on ADS-B monitoring server to generate reports as material for aircraft technician studies to make decisions about the feasibility of the aircraft on the next flight in preventing and reducing the rate of aircraft accidents.
I closed this research by implementing the model to the cloud system with ADS-B data and successfully identified its anomaly percentage. This research also has Intellectual Property Rights (HAKI) from the Ministry of Law and Human Rights of the Republic of Indonesia. If there is anything to ask or Indonesian aviation parties want to implement or modify the model, you can contact me directly on my LinkedIn (Abdul Azzam Ajhari) as a consultant.
How to cite my article (IEEE):
- A. Ajhari, & I. Negara “Aircraft Flight Movement Anomaly Detection using Automatic Dependent Surveillance-Broadcast,” JOIV : International Journal on Informatics Visualization, vol. 6, no. 4, , pp. 821–828, Dec. 2022. https://doi.org/10.30630/joiv.6.4.948
This article is a refinement of the articles we have made before and has many shortcomings:
How to cite my article (IEEE):
- A. A. Ajhari, R. Ibrahim, A. Pramodana, J. S. Pramudito, J. Reky Tasyam and W. Hilmy, “ADS-B Mobile Ground Station Receiver Flight Surveillance Architecture,” 2021 International Conference on Artificial Intelligence and Computer Science Technology (ICAICST), Yogyakarta, Indonesia, 2021, pp. 174–178, doi: 10.1109/ICAICST53116.2021.9497826.
- A. A. Ajhari, J. S. Pramudito and J. Reky Tasyam, “Rancangan Aplikasi Ads-B Pada Uav Dan Drone Komersil Dengan Raspberry Pi 3b,” Semnas Ristek (Seminar Nasional Riset dan Inovasi Teknologi), 2021, doi: doi.org/10.30998/semnasristek.v5i1.4787.
