Is a heavy PC no longer necessary? Secrets to lightening point cloud processing through cloud integration

Point cloud processing is a convenient technology that can record and analyze site conditions in 3D with high precision, but the reality is that along with its usefulness come challenges such as "heavy PCs" and "long processing times." Point cloud datasets that reach tens of millions to hundreds of millions of points can produce file sizes of tens of gigabytes, placing a heavy load on any computer just to handle them. In addition, field measurement work and in-office data processing and utilization tend to be disconnected, creating inefficiencies that prevent organizations from making immediate use of the data they have collected. This article delves into challenges in point cloud processing such as the PC specification problem, complex data management, and the gap between the field and the office, and provides a detailed introduction to workflow transformation via cloud adoption, an increasingly noteworthy solution. We explain the concrete benefits of cloud processing for lightening point cloud tasks, enabling division of labor and immediate sharing, and at the end of the article we present an operational example of the smartphone-mounted GNSS rover "LRTK" as a cutting-edge solution that enables field-complete, real-time processing.

Issue 1: The wall of file size and PC processing load

Point cloud data contains an enormous amount of information; scanning an entire site yields a collection of tens of millions to hundreds of millions of points. The data volume can easily reach several GB to several tens of GB per file. Naturally, handling such massive data requires commensurate machine specs. Loading large point clouds on a typical office PC causes severe sluggishness, and in some cases the software may freeze or crash. Comfortable point cloud processing essentially requires high-end CPUs, large amounts of RAM, and GPU-equipped high-performance workstations, so the hardware investment cost is not trivial.

Even if a high-performance PC is available, there remains the issue of processing time. Processes such as point cloud registration (merging), meshing, and analysis tend to take time proportional to the data volume, and in some cases a PC may run all night to complete processing. During such work, other tasks may be halted, reducing productivity. The excessive data weight and resulting lack of real-time capability are major hurdles to point cloud utilization.

As a stopgap, point clouds can be split by area or thinned by reducing point density to lighten the load. However, reducing the data in this way trades off loss of important detail and difficulty grasping the whole picture. Highly detailed point clouds that should be used to their full potential become wasted treasure if downsized due to PC limitations. As above, the problem that “the data is too large and processing is sluggish” is a wall that many engineers working on point cloud processing encounter.

Issue 2: Inefficiencies from the split between field and office

Another unavoidable issue in point cloud utilization is the workflow split between the field and in-house operations. The traditional flow is typically: measure and acquire point clouds in the field, bring the raw data back to the office for processing and analysis, and then produce deliverables such as drawings and reports. This “field → office → deliverable” sequence incurs a time lag, so even if 3D data is acquired on-site, it can take days to weeks before it becomes usable.

This split process creates various inefficiencies. For example, because data cannot be confirmed in the field, it is impossible to determine on the spot whether re-measurement or additional surveying is needed, and only after returning to the office and processing the data might one discover that “some parts were missed.” As a result, teams may have to return to the site for additional measurements, wasting time and incurring extra costs. There is also the hassle of transporting huge point cloud files via USB or external HDD or transferring them across the corporate network. Every time data is handed off among stakeholders, copies proliferate, and it is common to face confusion over “which version is the latest?” When different people have separate copies on their PCs and updates must be redistributed each time, digitized point cloud data cannot be fully leveraged.

Moreover, the split between the field and the office makes real-time collaboration difficult. Since the office cannot grasp field conditions on the spot, decision-making and directives suffer from time lags. For instance, confirmation of as-built condition or inspection approvals typically require field data acquisition followed by office-based analysis and checks, preventing immediate feedback on-site. If field personnel and office engineers could share data in real time, missed measurements or errors could be pointed out on the spot and additional measurements requested, but traditionally this was not possible, leading to the frustrating situation of “we can’t judge until we take it back and process it.”

As described above, the time loss and complicated information sharing caused by the disconnection between field work and subsequent processes in point cloud workflows represent issues with significant room for improvement.

The key to solutions: Revamp workflow through cloud adoption

One increasingly noted approach to solve the issues above is leveraging cloud services to transform workflows. Introducing cloud integration significantly changes how point cloud processing is handled.

First, it eliminates the need to hold massive point cloud datasets on local PCs. By uploading data to high-performance servers in the cloud via the Internet and performing heavy processing there, analysis proceeds quickly regardless of the specs of the user’s local PC. Users can check results in a browser or download only the necessary portions, meaning that, in extreme cases, even tablets or general-purpose laptops can handle large-scale point clouds. There is no need for each user to install expensive software or acquire the latest GPU-equipped machines; the cloud provides an always-up-to-date, high-speed processing environment.

Next, cloud mediation makes data management and sharing dramatically smoother. If point clouds acquired in the field are saved directly to cloud storage, all internal and external stakeholders can access the same unified dataset. When someone edits or adds measurements, those changes are immediately reflected in the cloud dataset, and other members can always view the latest information. There’s no worry about circulating an outdated version via email attachments or USB, and confusion about “which is the latest?” is resolved. For viewing and light analysis, users can work from a browser, eliminating the need to install heavy specialized software on everyone’s PC.

Critically, cloud integration enables real-time collaboration between field and office. With data on the cloud, the office can instantly view and analyze point clouds uploaded by field teams, and conversely, 3D design models created in the office can be overlaid and checked on a field tablet—enabling two-way interaction. Being able to communicate while viewing the same 3D data regardless of geography makes remote site inspections and rapid decision-making feasible. In short, cloud adoption addresses both the problems of point cloud “heaviness” and “slowness” while simultaneously solving the problem of “disconnection.”

What cloud processing can achieve

When point cloud data can be handled in the cloud, many previously cumbersome processes become faster and simpler. Below are representative functions and use cases enabled by cloud point cloud processing.

• Section generation: Instantly extract cross-sections at arbitrary positions from uploaded point clouds. You can check longitudinal and cross-sectional shapes on site or overlay profile drawings to check discrepancies. Cross-section verification that previously required extracting survey points on a PC can now be automated in the cloud.

• Heat map analysis: Overlay point cloud data and 3D design models to visualize deviations as color maps. For example, in as-built inspections, comparing post-construction ground point clouds with the design model and displaying a heat map where high areas are red and low areas are blue enables instant recognition of excesses, shortages, and unevenness. Because pass/fail judgments against tolerance can be shown intuitively by color, quality inspection efficiency is greatly improved.

• Quantity calculation: It is also possible to compute volumes and areas directly from point clouds in the cloud. For example, you can calculate fill and cut volumes for an earthworks area from point cloud data, or measure the length and diameter of scanned buried pipe models. Automatic aggregation of required cross-sectional areas and volumes makes it immediately useful for progress quantity calculation and estimating backfill volumes.

• 3D display and sharing: Large point clouds uploaded to the cloud can be displayed and manipulated smoothly in a browser-based 3D viewer. Displaying point clouds that used to be tens of GB was difficult, but cloud-side optimization and streaming enable smooth 3D viewing on typical PCs and tablets. You can share datasets by sending a URL link, and recipients can inspect 3D point clouds from any viewpoint without special software. This expands the scope of effective point cloud use for internal reviews, client explanations, and future maintenance data sharing.

• AR integration: Point clouds and model data in the cloud can be displayed as AR on field mobile devices. For example, overlaying cloud point cloud data on a tablet or smartphone camera feed synchronized with the real environment gives the experience of “seeing through” the point cloud on site. You can AR-visualize buried utilities from the road surface or overlay pre-construction design models on the site to visualize the finished state. This makes it easier to share realistic 3D images with non-technical staff and other departments, reducing communication loss and speeding decision-making.

What changes when field and office are connected?

When point cloud data is shareable in real time via the cloud, the boundary between field and office disappears, transforming how work gets done. Specific benefits include:

• Faster response through real-time sharing: Because the latest data acquired in the field can be immediately shared via the cloud, stakeholders can have reviewed the data by the time the field team returns. As a result, early detection of missed measurements or anomalies becomes possible, and field personnel can receive instructions for additional measurements or rework while still on site. This reduces unnecessary revisits after later data checks and shortens the project lead time.

• Immediate inspection and approval processes: As-built inspections, interim checks, and client approvals can be done in real time. Traditionally the flow was field measurement → office drafting → submission, but if the point cloud itself is shared in the cloud, authorized personnel can directly review and approve inspection data remotely. So-called remote site attendance becomes possible, allowing supervisory agencies or clients to audit the site 3D from their office and significantly reducing the days required for inspection and reporting tasks.

• Improved division of labor and parallel work: With centralized data in the cloud, multiple people can perform concurrent tasks. For example, while one person scans the site, another in the office can begin drafting and computing quantities from the incremental point clouds being uploaded. This overlap of processes blurs the boundary between fieldwork and desk work, allowing each specialist to proceed with their tasks based on real-time data and dramatically improving overall work efficiency.

Thus, by enabling seamless field-office connectivity through the cloud, time spent waiting for heavy processing, data transfer hassles, and post-field analysis delays are progressively eliminated. An environment where the whole team can collaborate in real time around shared point cloud data represents a new work style for the DX era.

Tips for acquiring point cloud data suited to cloud integration

To make the most of the cloud, it’s important to adopt methods that consider cloud integration from the data acquisition stage. How point clouds are measured can greatly affect downstream effort and the effectiveness of cloud utilization. The following are points for acquiring point clouds optimized for cloud workflows.

• Acquire point clouds with absolute coordinates: If you assign geodetic absolute coordinates to point clouds from the outset, post-processing becomes much easier. Typically, point clouds from terrestrial laser scanners or drone photogrammetry are recorded in various local coordinate systems, requiring later registration and alignment to known coordinates. But if each point has Earth coordinates from GNSS (GPS) or similar, all datasets are already aligned and can be handled as-is in the cloud. Point clouds whose positions are fixed in a global coordinate system make it smooth to compare with CAD design data or integrate with other site data, enabling immediate overlay and analysis in the cloud.

• Smartphone integration: Linking point cloud scanners or surveying instruments with smartphones is another key in the cloud era. Smartphones and tablets are increasingly becoming the platform for field work, and more devices now offer one-stop measurement-to-data-upload via dedicated apps. By pairing with a smartphone, acquired data can be uploaded to the cloud over 4G/LTE or Wi‑Fi on the spot, or cloud-based design data can be downloaded to the phone for field checks—enabling real-time synchronization between the field and the cloud. Modern smartphones also include LiDAR sensors and high-performance cameras, and for short distances a smartphone alone can acquire point clouds on the order of several million points. Combining these easy-to-use sensors with the cloud makes small-scale surveys entirely feasible with just a phone, allowing immediate cloud sharing.

• Surveying without control points: Omitting the setup of known control points during point cloud measurement can significantly reduce field time and effort. Traditionally, high-precision alignment required measuring ground control points with a total station or placing numerous target spheres for post-processing. However, advances in RTK-GNSS and networked correction services now enable centimeter-level positioning (half-inch accuracy) without dedicated local reference stations. Using these methods, high-precision coordinates can be directly assigned to acquired data without physically setting and surveying control points. Simplifying fieldwork by foregoing control point setups shortens the time to cloud upload and enables immediate analysis and sharing right after acquisition. This minimizes equipment setup and surveying personnel, enabling agile measurement operations optimized for the cloud.

These measures make the flow of “directly linking field-acquired point clouds to the cloud” smoother. Particularly, point cloud data acquired with absolute coordinates is highly compatible with cloud workflows, as it eliminates later alignment tasks and allows direct real-time processing and analysis.

Frontline of field-complete solutions: Example use of the smartphone GNSS rover "LRTK"

The fusion of cloud integration and point cloud surveying is already being realized in practical solutions. One representative example is the smartphone-mounted high-precision GNSS receiver system “LRTK.” LRTK is a palm-sized RTK-GNSS rover device that attaches to a smartphone, and combined with a commercially available iPhone it enables centimeter-level positioning (half-inch accuracy) for 3D surveying, representing an innovative device. The compact unit, weighing just about 165 g, mounts on the back of a smartphone and connects via Bluetooth, instantly transforming the phone into a high-precision surveying instrument.

A workflow using LRTK exemplifies the “acquire on-site → immediately share via cloud → real-time utilization” concept described earlier. For example, at a certain civil engineering site, simply walking through construction areas with a smartphone fitted with an LRTK captures surrounding terrain and structures as point clouds, and absolute coordinates are assigned to all acquired point clouds in real time. After measurement, you can upload data to the cloud with a single tap from the smartphone app, enabling instant sharing of results with stakeholders without returning to the office. Uploaded point clouds and coordinate measurement results are automatically organized and visualized on the dedicated LRTK Cloud (web app). Measurement points are plotted on a map, point clouds can be viewed in a 3D viewer, and photo- and note-attached reports for each point are auto-generated. Sharing the issued viewing URL allows recipients to freely view and download the data in a browser, so even remote supervisors or clients can check the site’s 3D data immediately without special software.

At the same time, field engineers can use the cloud-uploaded data to complete verification and reporting on the spot. For example, they can display cross-sections of uploaded point clouds on the smartphone and compare them with design sections, or check automatically computed as-built dimensions on the cloud to instantly judge compliance with construction standards (as-built inspection). For buried utility measurements, they can review backfill volume (quantity check) results calculated in the cloud and place additional orders on site if shortages are found. LRTK also integrates with AR functions so that acquired point clouds and 3D models can be overlaid on the smartphone camera feed for on-site AR verification. For instance, immediately after paving and backfilling, you can use the smartphone to see through and visualize the point cloud model of underground piping from the road surface—advanced verification that a single person can accomplish. Previously, surveying would be followed by office-based drafting and modeling, and viewing in special AR equipment—a time-consuming process—but with LRTK, a single smartphone on-site completes measurement, 3D modeling, and AR display.

This type of field-complete, cloud-integrated workflow using LRTK removes many barriers to point cloud processing. Without relying on heavy PCs or specialized software, anyone can acquire and use high-precision point cloud data in the field, making the question posed in the title—“Is a heavy PC no longer necessary?”—increasingly realistic. Being able to perform as-built inspections, quantity calculations, and visualizations immediately after data acquisition and share them with stakeholders on the spot dramatically improves both speed and accuracy. The lightening of point cloud processing and efficiency gains from cloud integration are not just time savings but a shift toward a field-centered new way of working.

To maximize the potential of point cloud technology, the key is to move beyond traditional heavy-handed methods and transition to smart workflows that incorporate the cloud and advanced devices. Fortunately, with cutting-edge tools like LRTK already available, the necessary technical foundation is falling into place. Field teams that hesitated to utilize point clouds due to data volume and PC load can adopt cloud-integrated solutions to leverage large datasets for real-time operational improvements. With a future of point cloud processing not bound by heavy PCs now within sight, this is an opportune time to incorporate the benefits of cloud adoption into your workflows and take the next step in point cloud data utilization.