News

Mission Impossible – 103 Quarries in under 4 days

We take a look back at a project we completed in the Summer of 2017 in Spain.

The Background

This project was a collaboration with Grup Air-Med, an aerial survey company near Barcelona. The aim was to see whether it was possible to use XCAM for an aerial survey project that was seemingly impossible within the time-frame. We like a challenge, so we got involved!

 

The Area

The project involved capturing airborne data for 103 individual quarry sites in high resolution. The majority of the quarries were not very big, with the smallest at 789 square metres (0.19 acres) and the largest at 2.33 square kilometres (575 acres). However, they were spread over a 5,000 square kilometre (1.2 million acre) area with some sites being on the periphery of towns, some amongst the bodegas and some being up in the mountains.

 

The Requirements

The client needed oblique images taken around each site, a Digital Surface Model (DSM) and a video flythrough around the site.

They also needed the data delivered within two months of being captured. These requirements throw a massive spanner in the works for traditional airborne sensors.

 

Choosing a Sensor

Large Format. A large format sensor could be used, especially one with both vertical and oblique sensors, but we would come across a number of challenges. Even for the small quarry sites a large survey block would be needed to get the obliques from all angles. This would result in a large amount of excess flying, adding cost and time to the project. We would also need to be flying at a much higher altitude to achieve the resolution required. Although not a problem in itself, it would limit the amount of flying that could be achieved in one day. Especially as in some of the mountainous areas they get a large amount of low-level cloud in the afternoon. This increases the time that a crew (and presumably a leased aircraft, pilot and sensor) are deployed for. The viability for this option therefore quickly falls away into an uneconomic mess.

UAV/Drones. Another option could be with a UAV. They are small and adaptable and can be set for flight plans to achieve the imagery we would need (possibly with a bit of customisation). However, for the majority of the sites a whole day would be needed to complete the capture, even with doubling up on the UAVs. This means that we’d be looking at over 90 consecutive suitable survey days. Of course, we could use multiple crews with multiple UAV, but there’s still no way the project could be even captured, let alone processed within the client timescale. The UAV also has a few other issues such as obtaining the permissions for flying over the sites. Many were over main highways, town centres and airports. Others were up in the mountains which, with the rapidly changing weather conditions, could cause huge operational issues. As with the large format camera system, the UAV presents too many risks for a commercial project.

LiDAR. A LiDAR system would capture the DSM very well but would require an additional survey to capture the oblique photos, so this was also discounted for being economically unviable and out of specification.

XCAM. After some detailed analysis and simulation of flight plans, (unsurprisingly for GeoXphere) the XCAM was chosen for the job. We’d be able to use a low-cost leased light aircraft combined with our XCAM Circle Path method of capture. We would be flying low enough to dodge the cloud as it appeared, we wouldn’t have any restrictions on where we could fly, and we could quickly navigate to the next site in a matter of minutes. It became a no-brainer so we moved ahead with the flight planning.

 

Flight Planning

For the flight planning we decided on the Circle Path as it would give us the flexibility to produce all the required datasets from one raw dataset; the oblique images. We grouped some of the smaller quarries into one flight plan so that for a small amount of flying (and later processing) we could tick off multiple sites.

The project sites were numbered based on the client’s own naming convention for the sites so that we could easily communicate progress and avoid unnecessary translations in the processing stage. We also accumulated the numbering so if one mission plan captured just one site then it would have one number, e.g. A1. If the mission plan captured multiple sites in one then it would have multiple numbers, e.g. B1_B2_B3_B4. This was not necessary operationally, but it created the accountability for making sure each site was ticked off. It also gave a feel-good factor once a bigger mission was completed!

A critical path analysis was also carried out to reduce the amount of decision making that would need to be done in the air. In a survey aircraft, a quiet radio is a happy radio. We knew that there would be some variation to this during the project, but it was useful to have a guide.

All the plans were loaded on the XCAM Tablet and a paper-based tables were printed off for the navigator to track progress and make notes as the survey progressed.

Aerial Survey Notes

 

Flight Operations

The joint GeoXphere / Grup Air-Med team factored in just 5 days for the complete acquisition phase. This included training the pilot in the Circle method of survey, capturing all the data, producing backups and de-mobilising.

XCAM installed on Cessna 172 in Spain

The team arrived at the local airport with a Cessna 172 and an XCAM B (dual RGB). The installation was slower than normal as we doubled it up as a training exercise for the pilot who hadn’t seen the system before, but in total took around 20 minutes.

The first quarry site, close to the airport was chosen as the training site for the survey. After about 2 minutes of wobbly circles the pilot got to grips with the display on the XCAM Tablet and from then-on he was capturing perfect circles. The site was re-flown and then the point-to-point capture began.

XCAM flying over quarries and vineyard

During the surveys we would try to keep to the critical path for maximum efficiency, but would deviate from this plan if there were clouds on the site at the time, or if we needed to break for fuel or lunch. This would be recorded in the paper-based flight sheet.

During the breaks the data was copied to a MicroSD card to ensure there was enough spare storage on the tablet to complete the second part of the day. At the end of each day the data would be transferred to portable disks and backed-up.

In total, the complete survey took just 3 and a half days. We would start around 9am and finish around 4pm with a 1 hour break at lunch. And we also learnt a few tricks so if we were to do it again we think we could shave maybe half a day off this.

Refueling for an XCAM survey

So that’s 103 quarry sites surveyed in 21 hours. An average of one site completed every 12.2 minutes. Not bad for a few days’ work!

 

The Results

Exactly 66,200 images were captured for the project. That’s 607 GB of data.

The resulting data from the processing was excellent using Photomesh. The 3D models were sharp and very detailed and the flythrough videos that were needed by the client looked almost photo-realistic.

We could go into detail, but the results speak for themselves. We’ve included some samples on our Data Showcase page, including a download for the TerraExplorer 3D viewer.

XCAM Quarries

XCAM Solar Panels

XCAM Quarries

 

Summary

In summary, XCAM was an enabler for this project. It would have been nearly impossible, and certainly not cost-effective, to use any other sensor. It also allowed the data to be delivered to the client in a third of the time it would have taken other methods to capture the data alone.

This method could be applied to other areas of business, including Estate management, forestry and agriculture. Basically, anything where there’s a need for survey data across multiple sites.

 

If you have an idea for a project like this, please contact the GeoXphere team. We’ll be able to put you in contact with a local XCAM operator, or we can help you use it yourself.