Q 1: What is 3D data? 3D data (point clouds/ LiDAR scans) has reliable depth, shape and location information, in addition to the colour and texture information available in 2D photos. This means that with reliable 3D data, you can take accurate measurements, understand what is a size of an element, how that element is related to other elements in the space and whether a specific element is behind or in front of other element(s) and by how much. As such, you need reliable 3D data to make sense of the world in 3D, for AR/VR applications, and to improve quality of photos.
Q2: Why do I need to collect 3D data in construction? for many reasons! the key one is that construction is a 3D physical environment changing on a daily basis. tTo “automate” critical processes such as progress tracking and installation quality verification, and to build semantic digital twins (not digital records), you need to have reliable 3D data, which would then be fed into intelligent AI engines for scalable automated analysis
Q3: Can’t I just collect the data using 360 cameras? 360 photos are 2D photos shown in 3D, not 3D data. Yes, you can use photogrammetry to stitch photos together and get a walkable, 360 visual record of your sites, but a) the underlying 3D data is very noisy and cannot be reliably used in automated AI pipelines (beyond applications related to safety, tracking equipment and people) b) 360 cameras are lighting dependent and very sensitive to clutter (the norm on construction sites), introducing significant performance limitations c) different photogrammetry methods have different results and the performance changes with the camera, clutter and lighting conditions!
|So, go with 360 cameras if your goal is a) to create a 360, visual documentation of your construction sites and b) to automate processes such as tracking people, equipment, safety, etc (in combination with 2D AI algorithms). But if you are after automated, scalable processes that can be systematically used across all your construction projects, turning site data into automated, actionable insights for the whole construction duration, and at scale, 2D Data and photogrammetry do not cut it, period.|
Q4: How can I reliably collect 3D data? currently through stationary LiDAR scanners (range varies depending on application and device), mobile LiDAR scanners and drones. Stationary LiDAR scanners have mm accuracy and mobile scanners are 10-20mm accurate. There are some semi-autonomous systems in the market, but while fully autonomous and small sensors are the future, they are not practical yet. The data collection device and mode should be determined based on the nature of your site (linear infrastructure vs buildings), its size, and the accuracy required (healthcare and transport projects require more accuracy compared to a residential property)
Q5: What is Apple LiDAR then? Apple has introduced the LiDAR technology to their iPhone 12 and latest iPad Pro lineups, but the current version cannot be reliably used in construction for global data collection given accuracy, range, etc. The current model is between a 360 camera and a LiDAR scanner and is developed with AR/VR applications, including interior scanning, human AR/VR, etc. in mind. At present in construction, you can use Apple’s LiDAR functionality to grab local LiDAR data for immediate, fast feedback that can be embedded into the reliable global 3D data. But do not take Apple LiDAR and similar technology lightly! a revolution is about to happen and soon 3D data collection will be an integral part of every construction project; so be prepped for it and build it into your strategy if you have not already done so.
Q6: I know 3D data is great, but the volume of the data is very big, what should I do? The challenges you have been experiencing is because you have been handling the data manually. You have some “pointed tools” that can help visualize Point cloud data and/or perhaps import parts of it into BIM models, etc. But these methods are slow, very manual and prone to error and they crash very regularly due to capacity constraints.
@ Contilio we have built smart algorithms that can structure the otherwise unstructured, meaningless point cloud data and filter it for what does not belong to the construction site, including trees, vegetation, material, scaffolding, etc. This helps reduce the data volume 20-fold, as well as to structure the point cloud data so it looks like your construction site. Our cutting-edge 3D AI engine (AI that works on top of 3D data and is similar to what is used in autonomous driving) automatically identifies and classifies every structural, architectural and MEP element in the Point Cloud and then compares the results automatically with the BIM and your schedule. This helps automate progress tracking, installation quality verification and predictive risk management for your construction projects, providing you with 20X value/ROI. We also create smart (semantic, structured and machine readable) digital twins, enabling you to realize further benefits during the whole asset’s life cycle. All the results can be accessed through a cloud-platform that is intuitive and easy to use. To learn more about how you could use Contilio for your construction projects and operating assets and the value you can create for your projects and organization, please reach out to us at email@example.com or book a demo with us on our website
For marketing and PR please contact Jack Li at firstname.lastname@example.org.