With continual advances in technology, the surveyors toolkit is changing rapidly and it’s difficult not to get caught up in the hype surrounding innovative new tools such as Drones.
While we acknowledge the importance of embracing new technologies, as a precursor to gaining efficiencies in our business and ultimately providing a better value proposition for our clients, it’s important we understand:-
- does photogrammetry produced from Drones, allow us to maintain the quality and accuracy of the data ?
- what are the ‘real’ efficiency gains, in utilising such technology ?
In order to answer these questions, we set out to test the Drone capabilities against our traditional Detail and Level Survey.
We picked a large Greenfield site, some 206 hectares in Sydney’s South West which included an operating chicken farm, existing housing, machine sheds, a river and an adjoining environmental fill centre.
Our strategy was to carry out the Detail & Level survey using GPS technology, together with the traditional Theodolite as required and then fly up to 30% of the property with the Drone. The Drone utilized for this project, supplied and operated by Bryce Dimec from C R Kennedy, was the DJI Phantom 4 with built in 12.3MP Camera. Drone specifications include:-
- 28 minute maximum flight time, maximum control range of 3.1mi (5km).
- Maximum speed of 44mph (72kmh).
- Optimized Vision Positioning System raises positioning altitude up to 10 meters
- The camera has a fixed focal length and is a digitised sensor and shutter allowing for no movement in mechanical parts during flight and utilising a gimbal for greater stability and smoother footage
- Built in GPS System with no RTK support
- Smart Return Home allows the Phantom to avoid obstacles as it returns home.
The comparison for detail and level was broken up into 3 parts, which included:-
- The survey for the Digital Terrain Model (DTM)
- The capture by Drone of many existing buildings and infrastructure
- Access to difficult areas along the riverbank
The traditional Detail and Level survey provided a satisfactory control network for the photography, to achieve an accuracy level of ± 100 mm, for the XY and Z co-ordinates. (In the absence of the control network, the Drone technology would only achieve an accuracy level of ± 500 mm, for the XY & Z co-ordinates). The open terrain on this particular project was ideal for Drone Capture, as the open paddock grass was almost non existent and therefore a direct comparison to ground levels, rather than top of vegetation, could be made.
Flying at a height of 30 meters above ground, the Drone surveyed approximately 60 hectares of land, which included all existing structures – housing, machine sheds, large chicken coups and a portion of the riverbank. It took approximately 6 hours in the field and a total of 2 hours flying time, where we collected some 2000 photo’s.
How Did It Go ?
The Drone data allowed us to detail all of the aboveground improvements including buildings, yards, roads, dam location, stock piles, fences, adjoining site information, tree & vegetation locations efficiently. We estimate this data capture using traditional field survey methods would have taken an additional 2-3 days.
A surprising outcome was the result of the river survey. We had expected the riverbank to be masked by vegetation, however it was well definable from the photography (flown at 50 meters), so together with the ground survey, allowed us to make an accurate determination of the river banks. The Drone capability certainly provides an ideal solution where there is difficult terrain or unsafe access to certain areas.
The Drone did prove to have a number of limitations. Similar to GPS survey, it wasn’t able to be utilized across areas of dense tree canopy or provide levels around areas of congested infrastructure, hence the area’s had to be surveyed by traditional means. Further, to provide the required accuracy in the areas of open terrain the photography required significant survey control and therefore these areas would be more efficiently surveyed by a mobile GPS platform.
The data processing capability is also a consideration with Drone technology. Processing of the flight data, some 2000 photo’s, is tedious and requires a well spec’d PC. Taking approximately 12 hours in total, the final file size of the completed project included 6GB of data and 10GB of photos. Outputs of such processing allowed us to provide Orthomosaic to our client, together with Point Cloud, Trueview and 3D Models.
The Wrap Up !
Overall, there was no real gain in efficiency across the course of the project. The need for rigorous survey control to provide required accuracy, coupled with the extended time taken for the reduction of survey data, well outweighed the reduced number of field days required for collection of data, that the Drone provided.
Until such time as the Drone can be fitted with scanning capability incorporating an on-board RTK system with a ground control station, accuracy will remain an issue.
As highlighted during the survey of the river, the real benefit with the Drone comes where data collection is limited by time or access is difficult or unsafe.
That being said, we cannot underestimate the ‘fun’ factor the Drone offers, as means for data collection. It most certainly has more appeal than measuring with traditional methods, particularly amongst the Gen Y employees.
We should also point out as a final consideration that a CASA issued UAV Controller Certificate is required in order to pilot an unmanned aircraft. Heavy penalties may be imposed and insurances negated in the absence of such a certificate.
Undeniably, the Drone most certainly provides some flair in the toolkit. We look forward to continued advances & improvements that will continue to evolve with this technology, which will allow its broader application.
Our special thanks to C R Kennedy for their involvement in this project.