- VIDEO: Find out how the DJI Zenmuse L2 is benefiting operations at the £550m Everton stadium development. Also view datasets and accuracy reports;
- On-site tests demonstrate the improvements of the L2, compared to its predecessor, the L1;
- Select Plant Hire - subsidiary of Laing O’Rourke - says the L2 will further enhance workflows, flying faster for greater efficiency and collecting more accurate data, even when deployed higher;
- Drone LiDAR has played an integral role in the development, including to monitor Dock infilling and for earthwork calculations;
- Photogrammetry and visual drone data also benefiting on-site operations;
- Drone data used in immersive suite alongside a digital 4D model of the stadium for real-time updates and realistic insights.
The DJI Zenmuse L2 LiDAR sensor is an accurate and efficient drone surveying solution - the team building the £550million Everton Stadium has said.
This robust point cloud shows how the L2 provides a detailed true-colour reconstruction of the 52,888-capacity arena, under-development at Bramley-Moore Dock, on the banks of the River Mersey, in Liverpool.
On-site tests - conducted by Select Plant Hire, subsidiary of main contractor, Laing O'Rourke - have verified the L2's enhanced accuracy: Achieving an absolute error of 37mm from 100m flight altitude, compared to 54mm with its predecessor, the L1.
Drone LiDAR has been used throughout the project, including for earthwork calculations and monitoring the infilling of the dock.
And the L2's key upgrades - including an increased return rate and higher sampling rate for better object detection - will further benefit the drone surveying team at Everton.
The DJI M300 Series and 45MP P1 camera, as well as the DJI Mavic 3 Enterprise, have been deployed on site for photogrammetry and visual imagery.
This data has enabled regular site monitoring, progress updates, and logistics coordination, and has been used in Laing O'Rourke's immersive suite to complement a 4D model of the stadium to visualise the construction sequence, assess potential risks or clashes, and optimise resource allocation.
Meanwhile, cinematic aerial content has helped keep fans and stakeholders updated on progress.
Arran Scallion, responsible for Select's UAV department, said: “Drones have played a key role in the Everton Stadium development, bringing a multitude of benefits for planning, forecasting and managing the construction process.
"As part of the development, we have used the L1 on-site, but having tested the L2, we can see that it will further enhance our workflows."
heliguy™ has supported the integration of drones on the Everton Stadium project, providing the hardware and delivering drone, survey-specific, and regulatory training.
Drone LiDAR On The Everton Stadium Development
Mr Scallion described drone LiDAR as a revolution for the construction industry, capturing 'vast amounts of data quickly and accurately' and producing precise point cloud representations of complex structures and environments.
LiDAR was particularly important during work to infill Bramley-Moore Dock, which was packed with 500,000 cubic metres of sand dredged from the Irish Sea to help create a solid base for the stadium.
The DJI L1 mapped the raising sand levels, allowing for precise planning and monitoring of the infilling process alongside project requirements.
The image below shows a dataset captured with the L1 and processed through DJI Terra during the infilling phase.
Mr Scallion said: “We needed real-time information to monitor the sand levels as they increased. LiDAR enabled us to collect this, quickly and accurately, on a daily basis and provide stakeholders with the key information."
Drone LiDAR has also helped to collect accurate stockpile information and monitor earthworks.
"We have generated data for earthwork movements and volumetrics," said Mr Scallion.
"This has been invaluable for a site of this size, obtaining these crucial insights more safely, efficiently, and cost-effectively than using traditional manual methods."
L2 Vs L1
While the L1 has been a mainstay of Select's fleet, the L2 will replace it, thanks to its key upgrades.
The table below shows the key specification differences between the two solutions.
DJI Zenmuse L2 | DJI Zenmuse L1 | |
Recommended Data Capture Speed | 15 m/s | 8 m/s to 13 m/s |
System Efficiency | The operating area of a single mission can reach 2.5km-squared. Relative altitude: 150m; Flight speed: 15m/s. |
The operating area of a single mission can reach 2km-squared. Relative altitude: 100m; Flight speed 13m/s. |
Detection Range | 450m @ 50% reflectivity, 0klx; 250m @ 10% reflectivity, 100klx |
450m @ 80% reflectivity, 0 klx; 190m @ 10% reflectivity, 100 klx |
Point Rate | Single return: max. 240,000 pts/s; Multiple returns: max. 1,200,000 pts/s |
Single return: max. 240,000 pts/s; Multiple return: max. 480,000 pts/s |
System Accuracy (DJI-stated specs) | Horizontal: 5cm @ 150m; Vertical: 4cm @ 150m. Both at 150m flight altitude, flight speed 15m/s |
Horizontal: 10cm @ 50m; Vertical: 5cm @ 50m. Both at 50m flight altitude, flight speed to 10 m/s. |
LiDAR Module: Ranging Accuracy | 2cm @ 150m | 3cm @ 100m |
Inertial Navigation System Accuracy | Yaw Accuracy: Real-time 0.2°, post-processing, 0.05°; Pitch/Roll Accuracy: Real-time 0.05°, post-processing, 0.025° |
Yaw Accuracy: Real-time 0.3°, post-processing, 0.15°; Pitch/Roll Accuracy: Real-time 0.05°, post-processing, 0.025° |
LiDAR: Maximum Returns Supported | 5 | 3 |
LiDAR: Maximum Sampling Frequency | 240 kHz for all modes, including Penta | 240 kHz (single/dual echo mode); 160 kHz (triple-echo mode) |
LiDAR: Scan Modes (including FOV) | Repetitive scanning pattern: 70° x 3°; Non-repetitive scanning pattern: 70° x 75° |
Repetitive scanning pattern: 70.4° x 4.5°; Non-repetitive scanning pattern: 70.4° x 77.2 |
RGB Mapping Camera | 4/3 inch; 20MP; 0.7 seconds shooting interval |
1 inch; 20MP; 2 seconds (minimum) shooting interval |
The L2's enhanced capabilities lead to:
- Increased return rate and higher sampling rate, improving he L2's ability to detect small details.
- Increased efficiency - fly further and higher, plus the L2 does not require a 5-10 minute period for the IMU to warm up.
- Denser point clouds, thanks to the L2's more concentrated beam. The L2 has a reduced spot size of 4 x 12 cm @100m, only a fifth of that of L1.
- 30% increase in the measuring range.
- Enhanced and more realistic datasets - collected more efficiently, thanks to the L2's larger RGB mapping sensor and faster shooting interval. Generate true-colour point clouds and deploy for photogrammetry.
To compare the two solutions, the L2 and the L1 were flown over the Everton Stadium development, using a DJI M350 RTK drone. The following types of data was collected:
- Nadir, with the camera pointing downwards;
- Oblique, to map the façades;
- Manual scanning around features of interest.
L2 vs L1 Datasets
The comparison datasets - as displayed in DJI Terra - show some differences in the resulting outputs.
The L2 point cloud contains greater detail and clarity: The roof structure is more defined, the seating area is more complete, and the cranes have more shape.
This is a result of the L2's enhanced sampling rate and more concentrated beam, enabling it to better detect smaller and more intricate details.
heliguy™'s GIS lead said: "The L2 has upgraded visualisation features and there are some clear improvements in the data quality when comparing the L2 against the L1. The data is cleaner, with very little noise."
After processing through DJI Terra, the data can be input into third-party software, like Terrasolid, for further manipulation.
For instance, this image below shows how the dataset has been classified in Terrasolid, based on criteria such as ground, walls, roof etc. The platform's wizard conducts this process automatically, but this can be done manually for further clean-up.
And these images show how the data can be transformed into a Digital Surface Model...
...and a Digital Terrain Model, stripped down to just the ground surface.
DJI L2 vs L1 Accuracy
And what of the accuracy?
In the comparative flights at 120m, the results of the L2 surpassed the L1.
This is demonstrated in the two accuracy tables below: The first shows the results for the L2 while the second displays the outputs from the L1. 10 control points were used in both flights.
L2 Accuracy Table
GCP | Easting | Northing | Known Z | Laser Z | Dz |
1 | 333354.589 | 392605.931 | 6.748 | 6.733 | -0.015 |
2 | 333661.741 | 392595.724 | 7.305 | 7.301 | -0.004 |
3 | 333340.232 | 392595.667 | 6.876 | 6.899 | +0.023 |
4 | 333245.281 | 392454.656 | 6.620 | 6.597 | -0.023 |
5 | 333309.760 | 392379.181 | 6.345 | 6.332 | -0.013 |
6 | 333578.549 | 392291.515 | 6.979 | 7.005 | +0.026 |
7 | 333437.283 | 392237.129 | 6.720 | 6.769 | +0.049 |
8 | 333379.285 | 392221.457 | 6.950 | 6.977 | +0.027 |
9 | 333582.585 | 392211.304 | 7.368 | 7.218 | -0.150 |
10 | 333366.341 | 392195.183 | 7.114 | 7.153 | +0.039 |
Average Dz: -0.004 Absolute Error: 0.037 |
L1 Accuracy Table
GCP | Easting | Northing | Known Z | Laser Z | Dz |
1 | 333354.589 | 392605.931 | 6.748 | 6.762 | +0.014 |
2 | 333661.741 | 392595.724 | 7.305 | 7.358 | +0.053 |
3 | 333340.232 | 392595.667 | 6.876 | 6.946 | +0.070 |
4 | 333245.281 | 392454.656 | 6.620 | 6.614 | -0.006 |
5 | 333309.760 | 392379.181 | 6.345 | 6.414 | +0.069 |
6 | 333578.549 | 392291.515 | 6.979 | 7.034 | +0.055 |
7 | 333437.283 | 392237.129 | 6.720 | 6.822 | +0.102 |
8 | 333379.285 | 392221.457 | 6.950 | 7.014 | +0.064 |
9 | 333582.585 | 392211.304 | 7.368 | 7.386 | +0.018 |
10 | 333366.341 | 392195.183 | 7.114 | 7.207 | +0.093 |
Average Dz: +0.053 Absolute Error: 0.054 |
The results show that the L2 achieved enhanced vertical accuracy, compared to the L1.
It means that flights can be conducted higher, resulting in greater efficiencies in the process. The L2 is able to fly at a faster speed (15m/s), compared to the L1 (recommended between 8m/s and 12m/s) to further reduce surveying times.
Mr Scallion said: "The improvements of the sensor mean you can fly faster for greater efficiency and collect more accurate data, even when flying higher.
"Increasing flight paths not only increases the speed of the data collection, but it also means we can fly further away from people, improving safety in the process.
"The improved RGB camera alongside the LiDAR module means that we can collect LiDAR and high-resolution imagery at the same time - and see this side-by-side during a mission - for more robust and accurate datasets.
"The subsequent data can be processed quickly through DJI Terra."
For the Everton Stadium team, the upgraded L2 brings a range of benefits and will further aid LiDAR and photogrammetry data collection workflows.
Mr Scallion said: "heliguy™ has been very supportive in directing us into the latest technologies and helping us understand the benefits of the various solutions on offer from DJI.
"LiDAR provides a comprehensive representation of the construction site, with precise dimensions, and offers valuable insights to architects, engineers and project managers, ensuring a solid foundation for project planning and design.
"It expedites data acquisition and analysis, facilitating better decision-making and reducing design errors, allowing construction projects to stay on schedule and budget and ensuring the highest quality standards are maintained throughout the project."
Using Drone Data To Complement Digital Twin
Drone LiDAR data has played an integral role, but other methods of drone data have also been pivotal throughout the stadium's development.
Drones have been deployed to monitor progress and to chart the evolution of the build - as these images show.
Drone data has also been used in conjunction with a digital 4D construction model of the stadium built on third-party software platform, Synchro.
The model is linked to detailed construction programme information and true-to-life, real-time drone data.
It allows the Laing O'Rourke team to monitor and manage the project, assess potential risks or clashes in the programme, and identify efficiency opportunities that might not have been seen using more traditional methods.
The rolling model can be accessed by the workforce on site, via phones and tablets, and takes away the risk and provides certainty on the design.
Gareth Jacques, Project Director for Laing O’Rourke at Everton Stadium, said: “Drone technology has been used extensively on this project.
"As well as doing normal progress videos and updates, we have used it in our immersive suite, which is a digital area where we coordinate construction work, logistics, plant movements, and cranage.
"We can look at our digital 4D model, our construction programme in its traditional format, and then we bring the drone technology in to see where we are in terms of reality. It gives us a complete spectrum in terms of theory, model, reality, and it has proved to be extremely effective.
"It has helped our construction management processes by enabling us to visualise the construction sequence, assess the potential risks or clashes in the programme, and improve the collaboration between the supply chain and our direct contractors.
“We have the information at our fingertips and it is a daily tool as part of the management process."
Drones have been useful for external communications and fan updates, as shown in the video below which celebrates the installation of the ground's turnstiles.
Mr Jacques said: “Drone footage has been vital for updating the fanbase and clients. It has helped communicate what we are going to deliver, or what has been achieved, and provides a source of reality compared to 2D drawings.”
The stadium - the largest single-site private sector development under construction in the UK - is scheduled to be completed before the start of the 2024/25 football season.
It will act as a catalyst for more than £650m of wider regeneration in North Liverpool - including Liverpool Waters and the Ten Streets Developments.
Once complete, the arena will attract more than 1.4 million extra visitors to the city each year, injecting tens of millions of pound into the local economy through retail, tourism and hotel occupancy.
To discuss any of the content in this article, including adding the L2 to your workflows or finding out how the heliguy™ in-house surveying department can help you start and scale your drone programme, contact us.
Leave a comment