#004- Why is it dangerous to use Pix4DCatch?
Where is heading the reality capture industry?
Lately, I read about new software from Pix4D S.A. to utilize Apple’s “LiDAR” sensor. In combination with a viDoc RTK (Real Time Kinematic) rover, a user is able to do reality capturing of the limited area with RTK precision and accuracy of the combination of ‘LiDAR’ and photogrammetry. Output data are georeferenced point cloud and mesh that can be created in Pix4DMapper or Pic4DCloud. Users can share it with customers or coworkers as files respective to content or as a link to Pix4DCloud. Click a link below to open 3D viewer.
Advanced photogrammetry software that uses images to create professional orthomosaics, point clouds, models and more.
It is advertised as the perfect solution for non-experts in reality capturing. On Pix4D’s blog, we can understand that this app is good to:
- send data about the site to clients without unnecessary site visits
- show stakeholders the exact dimensions of the pit, as well as the status of the utility being inspected and the work needed
I can see benefits in some use cases but I believe it is more dangerous for projects when it falls in the wrong hands.
Since Apple released the iPad with the ‘LiDAR’ sensor onboard, many different experts in their fields began interesting in using Apple’s scanner in their daily work. I can see the tendency to create as simple as possible devices with associated software that will create and deliver ‘perfect’ data for the end-user. As an example can be used Matterport, BLK2GO, or BLK360. These are presented as very simple but powerful devices, and they are, no doubt about it. The case is that they are not dedicated to people who don’t have any knowledge and understanding of those technologies.
What is the problem to hold BLK2GO stable during the walk around the building, upload the collected data to Cyclone Register360 and export a point cloud, could say someone? Then suddenly pop up issues because the data is misaligned or it supposes to be aligned with data from other sensors. The project gets stuck. The company loses money and time to solve the problem, and people get stressed because of the situation. Wouldn’t be easier to hire someone who has education or at least experience in reality capture?
I met many people and companies who would like to build or expand their businesses with reality capturing. They are architects, developers, water supply and sewerage sector’s companies, HVAC engineers who saw the commercial how simple is to create digital twin — a very popular word lately, based on laser scanning or photogrammetry. Those companies buy equipment for thousand of dollars but they either are not able to achieve satisfying 3D data or they use inaccurate 3D data in their daily work (with or without awareness of the quality). In most cases, they finally fail or put a lot of effort into it with unexpectedly, for them, weak results.
Why does it happen? The answer is simple! Not everyone can do it even though producers of software or hardware advertise their products as useable for everyone. Why not? Some of the people don’t have spatial imagination and literally do not see in three dimensions on a screen. Others are not willing to learn the basics of technology and principles which they try to use. They don’t understand what went wrong and why the result is bad.
Apple’s ‘LiDAR’ made even more mischief than other easy-to-use devices. As far as I remember it supposed to be a toy for simple scanning of small stuff for 3D printing or supplementation for an AR-based app. Some clever brands found out it can be also used in a professional job, which is great! Unfortunately in most scenarios, those solutions are used in inappropriate ways or for purposes it was not invented for.
Let’s go back to Pix4DCatch app. I didn’t use it yet and I will not write anything about the simplicity of the collecting and processing data workflow. I rather want to point out scenarios of how it would be incorrectly used. On the official website, this app is described to use for scanning groundworks, so I will use that to present possible scenarios with the usage of Pix4DCatch with iPhone, and viDoc.
Company X does groundwork and lays water/sewerage pipes in the ground, scans excavation, and then fills it in. It happens before a landsurveyor would come to measure pipes or even without calling a landsurveyor. Why does it happen? It is because very often the excavation is relatively small, which doesn’t give a big income for X, but the waiting time for a landsurveyor generates a loss. The excavator is not used then and does not earn money for company X. Then X sends the data to the ladnsurveyor and expects to extract all of the information from the point cloud. Unfortunately, our landsurveyor doesn’t know are the data correct, what is the accuracy of GNSS position, and the measurement itself. X claims that they did their job and even delivered measurements. X’s employee does not know what is multipathing, which side of the sky should be open to getting better GNSS measurement’s precision and the pictures cannot be blurry. The landsurveyor has to take responsibility for the mapped information — position in XYZ of the pipes and all of the related elements and the height of the surrounded ground. In this scenario, he has two options, either accept the X’s measurement and hope that the quality is good enough, or discuss with company X, listens to its complaints, and do his job properly. My quite long experience as a landsurveyor on construction sites suggests me this scenario would happen too often.
An employee from company X, who does the scanning, does not understand the difference between coordinate systems so scans in any coordinate system or in the last used coordinate system. What if the current project has a requirement to maintain it in a different coordinate system? The landsurveyor has to transform the data to the correct coordinate system which in case of long excavation and different cartographic projections in those coordinate systems would create extra inaccuracy in the data.
Those are only two scenarios that popped up in my head but I believe could be found more. I believe the app with RTK antenna for iPhone/iPad is really great solution as supplementation of landsurveyors/geomatics daily work or at least good trained employee of a company from other industry than the surveying. I understand that producers of surveying equipment or software want to sell their products as much as possible even though it would be done for inappropriate people, but it can make more mischief than good. I think also that people to easily believed in Apple’s devices as scanners and trust too much in data quality it provides.