Using Unmanned Aerial Vehicles (UAV) for mornitoring the construction progress of Ho Chi Minh City urban railway project

This paper describes a new approach for monitoring the construction progress of the Urban Railway Construction Project “Metro line1 Ben Thanh Suoi Tien” by using Unmanned Aerial Vehicles (UAV) to capture high resolution imagery at different stages of the project. The advantage of the AscTec Falcon 8 systems lies in their high flexibility and efficiency in capturing the surface of an area from a low flight altitude. In addition, further information such as orthoimages, elevation models and 3D objects can easily be processed by Pix4Dmapper software. The Ground Control Points (GCPs) and GIS data were used to compare the achieved accuracy of UAV method. This study shows the feasibility of using an UAV system for acquiring the high resolution aerial images and the new opportunities for managing construction progress over time.


INTRODUCTION
To develop remote imaging techniques that is able to provide high resolution, real-time images for social applications and economic development, as well as for national security applications is important items of Viet Nam's strategy in space research and applications until 2020 officially approved by the Prime Minister in June 2006, according to the Decision number 137/2006/QĐ-TTg. However, Viet Nam has been mainly using satellite images (taken from 500km to 900km altitudes) and images taken from aircraft for altitudes from 300m to 1.000m. The high-resolution imaging system based on UAVs being capable of taking images of low altitudes ranging from 10m to 300m is highly demanded and promisingly applied in many  [1][2][3][4][5][6] showed that the shortcomings of above-mentioned imaging techniques can be overcome by the remote imaging system based upon unmanned air vehicle (UAV).
This study presents the experimental results of UAV for monitoring the construction progress of the Urban Railway Construction Project "Metro line1 -Ben Thanh -Suoi Tien". The paper also contributes a new way to build up high resolution images for small and constrained areas in real time. Therefore, the use of UAVs is an opportunity for surveying that will be used where a need of high accuracy is required and fast data capturing is demanded.

Test Areas
To monitoring the construction process, the aerial images of the 20.9 ha Depot area (20.9 ha) and the construction line (17.2 km) are captured.
The imaging frequency is every 3 months. All of images are referenced to Coordinate Systems VN2000 (See figure 1).

Figure 1.
Overview image of the site to take aerial photographs and two small areas of high resolution images were taken from camera: Sony NEX-5N

Used Systems for the UAV Method
The UAV is developed using a commercially available, namely called AscTec Falcon 8 of which 3 main components are a remote control, the flight control software and the octocopter itself ( Figure 2). Camera Sony NEX-5N (16.1 megapixel) is used for remote imaging system actively with stabilized camera mount. The Falcon 8 major technical specifications include: Flight time: 20 -30 minutes (limited by battery capacity); Image transfer frequency (5.8 GHz); Control signal frequency from ground station (2.4 GHz) and Live video feed to the Mobile Ground Station.
The UAV method for the acquisition of geodata is based on a good and appropriate flight planning. In Figure 3 the flight planning for the metro line 1 is shown, using the provided software of the UAV manufacturer (AscTec Auto Pilot Control -Ascending Technologies).  The UAV has been experimented for the observed area and results showed that the dependence of horizontal resolution as function of flight altitudes.
From experimental results, it was necessary to set up a reference station during data acquisition by UAV for monitoring construction progress of the Metro line. With the help of these flight plans the UAV is steered autonomously over predefined routes. Along these paths, aerial images are taken at 100m for altitude, 65% overlap along track (flight direction) and 50% overlap across track. The Falcon 8 could be planned to operate in two different modes: (1) autonomously flight and take images at prescribed GPS-defined locations (2) manually flight control (real-time images observed) and take images at preferred locations. In order to transform the acquired data into the national coordinate reference system (VN2000), control points from official topographical surveying are required. To determine the relationship between altitude capture and image resolution, we conducted a flight test taken at different heights (30 m, 100 m and 150 m). The test results are shown in Table 1.

RESULTS
The mission (flight and data acquisition) is planned in the lab with AscTec Auto Pilot Control software. Thus fixing the image scale and camera focal length, the flying height is derived. The take-off and landing operations are strictly related to the characteristics of Falcon 8 and functions of remote controller. During flight, the data link, camera control, video link, status display and the controls for the system itself are integrated into the Mobile Ground Station.        Feasibility of using an UAV system for managing construction progress over time is showed by Figure 7. The necessary information related to Thao Dien station and scope of works to be showed by progress photographs of UAV that is carried out once in every 3 months during construction progress of metro line.
In addition, Pix4Dmapper image processing software can permit to create DSM model and 3D images for large scale area of metro line based upon overlapped images.

CONCLUSIONS
The achieved accuracy of data acquisition by UAV for monitoring construction progress of the Metro line depends on the flight height and have a 2D coordinate quality of 2.9 cm (image resolution) when the height of UAV is 100 m.
The optimal height of capture is 100m. The accuracy can be increased by using additional information of data from Ground Control Points (GCPs). This problem will be define in next our research when GCP network of construction is finished. Experimental results are very promising and clearly demonstrated the capacity of UAV-based remote imaging system to provide high-resolution image (with resolution lower than 3cm in horizontal plane and lower than 20cm in altitude) for for managing construction progress over time with quite low cost. This opens a new way that is suitable to be used in addition to the standard surveying methods in order to gain further data through the acquired images such as overview images or ortho-images and 3D models.