Six Precautions and Workflow to Avoid Failure in Creating 3D Data for Cultural Heritage

Creating 3D data is becoming part of routine fieldwork for the preservation, investigation, and public use of cultural heritage—not a specialized skill reserved for a few. Until now, records of cultural heritage have centered on photographs, measured drawings, and written descriptions. However, considering factors such as complex shapes, tracking changes over time, before-and-after comparisons for repairs, remote sharing, and reuse for exhibitions and education, two-dimensional records are increasingly insufficient. That is why 3D data, which can preserve the form itself in three dimensions, is attracting attention.

At the same time, creating 3D data for cultural heritage is not simply a matter of measuring an object and producing point clouds or models. If you proceed without understanding the value of the object, you may fail to achieve the necessary accuracy. Misreading site conditions can lead to missing or distorted data. Starting work without clarifying deliverable formats can leave you with data that is unusable later. In other words, what really matters in creating 3D data for cultural heritage is not the names of equipment or software, but the ability to design the project by working backward from the objective to eliminate potential failure causes.

What is especially troublesome for practitioners is that multiple purposes—preservation, research, management, publicity, and education—often coexist in a single project. Stakeholders expect different deliverables, and available on-site time is limited. It is not uncommon to face restrictions such as not being allowed to touch the object, inability to erect scaffolding, unstable lighting, or limitations on bringing equipment in. As a result, even when measurements are technically feasible, the project can still be close to failure from an operational perspective.

Therefore, pre-planning and preparation strongly influence outcomes in cultural heritage 3D data creation. If you can first clarify what you want to preserve, what level of accuracy is required, who will use the data and how, and how it will be reused in the future, on-site decision-making becomes much easier. Conversely, if these points remain vague, you will incur retakes, revisits, reediting, and re-deliveries, greatly increasing time and effort.

This article organizes six precautions that practitioners should keep in mind to avoid failure when creating 3D data for cultural heritage, and explains step-by-step how to proceed. It is written from a practical perspective to address common on-site pitfalls, useful both for those approaching 3D data creation for the first time and for those with some experience commissioning or handling fieldwork.

Table of contents

• Why failures are likely in creating 3D data for cultural heritage

• Precaution 1: Define the purpose and intended uses first

• Precaution 2: Don’t leave required accuracy ambiguous

• Precaution 3: Don’t underestimate site conditions and object characteristics

• Precaution 4: Plan acquisitions assuming missing data and noise

• Precaution 5: Decide deliverable formats and management methods in advance

• Precaution 6: Plan for ongoing operation and updates

• How to proceed to avoid failure in cultural heritage 3D data creation

• Conclusion

Why failures are likely in creating 3D data for cultural heritage

Failures in creating 3D data for cultural heritage often arise more from misaligned project design than from a lack of measurement technology. This is because cultural heritage differs from typical structures or construction targets: preservation constraints, usage policies, and recording priorities vary greatly by object. For example, some projects only need the external shape, while others require preservation of fine surface undulations or traces of damage. Whether the goal is capturing the overall shape, comparing deformations, or creating exhibition content drastically changes how the data should be produced.

Furthermore, cultural heritage projects often do not conclude with a single activity. Later uses such as investigation reports, repair design, conservation records, exhibition production, research sharing, or educational use may be added. Therefore, if specifications are determined solely for immediate use, the data may become hard to reuse months or years later. Data that has been overly reduced in size or lacks proper georeferencing tends to be especially limiting for future reuse.

Additionally, field conditions for cultural heritage are not always measurement-friendly. Sites can be narrow, dark, have elevation differences, be inaccessible, contain many obstacles nearby, be prone to reflections or shadows, or be affected by outdoor weather. Even objects that look simple may have many parts that are easy to miss, and on-site misjudgments easily lead to data gaps. Areas especially prone to omission include object backsides, eaves, stepped sections, deeply recessed ornamental parts, and faces close to vegetation or fences—places that are easily missed without planning.

Also, differences in stakeholders’ perspectives can cause failure. Preservation staff emphasize maintaining original form, researchers emphasize shape accuracy, and publicity teams emphasize appearance and clarity. Each stance is valid, so satisfying only one does not mean the project is a success overall. Therefore, practitioners must design the project not only around the measurement work itself but also around clarifying objectives, defining deliverables, and sharing operational constraints.

To succeed in creating 3D data for cultural heritage, it is essential to verbalize what constitutes success before choosing technologies. Whether this foundation exists or not will change everything in the subsequent fieldwork, editing, delivery, and use.

Precaution 1: Define the purpose and intended uses first

The first precaution is to make clear from the outset what the 3D data is being created for. This seems obvious, but in practice it is the part that tends to be most ambiguous. The phrase “3D data for cultural heritage” alone does not determine required specifications. Preservation records, condition assessment, repair planning, shape comparisons, exhibition use, educational materials, and public archiving all require different levels of detail and different deliverables.

For example, if the main purpose is preservation records, stable reproducibility for future comparison is more important than visual attractiveness. If you expect to compare pre- and post-repair differences, you need to organize conditions that make re-acquisition consistent. If you anticipate exhibition use, accessibility, lightweight models, and visually clear presentation are required. For research, retaining raw preprocessed data and alignment with coordinate systems matters. As the purpose changes, so does the definition of good data.

If this is left vague, on-site staff cannot decide priorities. Should you broadly capture the overall shape, densely record some parts, prioritize color information, or enforce strict positional references? Without such decisions, the result will often be half-baked. Something may look impressive visually but be unusable later; that state stems from insufficient early clarification.

As a practical step, list expected use cases concretely. Who will use the data, when, and for what decisions? Is the focus on viewing or also on measurement? Is this a single-year deliverable or the basis for ongoing management? Is it for internal use only or for public release as well? Clarifying these points makes required resolution, acquisition scope, editing policy, and deliverable formats much easier to determine.

Also, in cultural heritage projects it is important not to try to meet all uses perfectly with a single dataset. Thinking in terms of source data, analysis-ready data, and viewing-friendly data by use case makes realistic design easier. If you separate uses from the start, you can avoid on-site acquisition shortfalls while still making it easier to derive necessary derivative datasets later in the workflow.

Purpose definition is not mere preliminary explanation; it is the decision basis to avoid confusion in later stages. If you want to succeed in cultural heritage 3D data creation, start by clarifying uses rather than jumping into technical discussions.

Precaution 2: Don’t leave required accuracy ambiguous

The second precaution is not to leave required accuracy ambiguous. The term “accuracy” is commonly used in creating 3D data for cultural heritage, but its meaning often differs among stakeholders. You must separate concepts such as acceptable positional agreement, the level of surface detail to reproduce, whether overall consistency is important, and whether some fine local shapes are critical.

A common mistake is to define specifications with the vague term “high accuracy.” However, “high accuracy” alone does not enable fieldwork or editing. The workflow differs entirely depending on whether you mean overall positional registration accuracy, surface reproduction detail, or cross-section-level precision. Cultural heritage objects often have complex shapes, irregular ornamentation, and damages, so defining accuracy concretely is even more necessary than for typical structures.

For example, if the goal is recording the overall shape, continuity of surfaces and minimal distortion are prioritized. If the goal is comparing damage or weathering, reproducibility of local surface undulations is critical. If you plan to integrate with coordinate-referenced records in the future, you also need to preserve positional references, not only relative shape. If these purposes are confused, it becomes impossible to evaluate whether the deliverables meet requirements.

Practically, link required accuracy to use cases rather than stating numbers alone. Specify what needs to be readable, what magnitude of differences you want to compare, whether the data will be used for drawing or quantity assessment, or if it is viewing-focused. Defining the specification from the viewpoint of what decisions the data will inform reduces recognition gaps with the field.

Furthermore, required accuracy for cultural heritage may not be uniform across the object. It can be effective to capture the overall object at a standard density while recording ornamented or damaged parts at higher density. Attempting to acquire everything at the same density increases workload and can dilute effort on truly important parts. Rather than uniform accuracy, allocate the necessary density where it matters.

Clarifying required accuracy early improves not only quality assurance but also operational efficiency. To reduce the burden of over-acquisition and ensure essential records, concretize accuracy as soon as possible.

Precaution 3: Don’t underestimate site conditions and object characteristics

The third precaution is not to underestimate site conditions and the object’s characteristics. In cultural heritage 3D data creation, attention tends to focus on the object’s value and shape, but the real determinants of success are the constraints you face on site. No matter how good the plan, if you misread access paths, surrounding environment, weather, illumination, obstacles, or work permit conditions, unexpected issues will arise during fieldwork.

You may not be free to move around at a cultural heritage site. Access areas may be restricted, and touching or installing equipment may be prohibited. There may be lighting restrictions and limited working hours. You might have to work in short windows due to opening hours or visitor flows. Outdoors, sunlight, wind, rain, moving shadows, and nearby traffic can all interfere. Even indoors, narrow spaces, low light, reflective surfaces, and restricted footing can prevent work from proceeding as assumed on paper.

Pay attention to object-specific characteristics as well. Cultural heritage objects often include complex sculptures, deep shadows, thin members, missing parts, degraded sections, warping, and overlapping shapes—many elements that are easy to miss. Surface conditions that look straightforward can produce noisy data during acquisition. Differences in color or material can affect acquisition stability and the difficulty of post-processing.

For these reasons, you should not skip site inspections or preliminary interviews. At minimum, confirm the object’s approximate dimensions, surrounding space, directions from which you can approach, elevation differences, presence of obstacles, available working hours, availability of power and rest space, access routes for equipment, and provisions for rainy conditions. If a site visit is difficult, use drawings, photos, past records, and explanations from site staff to concretize conditions as much as possible.

Also, ensure that measurement itself does not become an end in itself. Avoid placing load on the object, adhere to management rules, and ensure your work does not inconvenience users or the surrounding environment—these are equally important. Not underestimating site conditions is not only about measuring efficiently but is a basic stance to balance preserving cultural value and on-site operations.

To avoid panic on site, design not only what to measure but also the environment in which you will measure. Creating 3D data for cultural heritage is as much about logistics as it is about technology.

Precaution 4: Plan acquisitions assuming missing data and noise

The fourth precaution is to plan acquisitions assuming that missing data and noise will inevitably occur. The most dangerous assumption is that you can capture everything perfectly in a single pass. Even when things seem visible on site, checking the data later often reveals omissions in blind spots, overlapping regions, recessed areas, and deep cavities. In addition, people, vegetation, fences, or background elements may be captured, and noise cleanup can take more time than expected.

Cultural heritage objects often have deep relief, fine components, and intricate surface ornamentation, so broad, coarse coverage is insufficient for adequate records. Parts that are visible head-on may not be fully defined unless captured from oblique angles. It is necessary to first capture the whole and then perform additional acquisitions for important or shadow-prone areas.

What is important here is to decide in advance what to check on site. Anticipate where omissions are likely, where it is hard to revisit later, and which parts are difficult to supplement afterward. Doing so reduces oversights during fieldwork. For example, treat backsides, bases, joints, under-eaves, deeply recessed decorations, stepped areas, and faces near nearby obstacles as priority checkpoints.

Do not skip immediate on-site verification after acquisition. Relying on post-processing to fix issues is particularly risky in cultural heritage projects. At sites that are hard to revisit, whether you can spot deficiencies on the spot is critical. Check for overall gaps, local omissions, positional shifts, and unwanted inclusions, and include time in the schedule for supplemental acquisitions when necessary.

Regarding noise, adopt a mindset of reducing noise in the field rather than depending on post-processing. Knowing potential noise sources—pedestrian traffic, wind-blown moving elements, unwanted background, backlight or strong shadows, and wet surfaces—can substantially change downstream workload. Slightly improving acquisition conditions on site often significantly reduces editing labor.

In cultural heritage 3D data creation, avoiding omissions is more important than obtaining pristine captures. Planning from the outset with the assumption that omissions and noise will occur leads to consistent practical results.

Precaution 5: Decide deliverable formats and management methods in advance

The fifth precaution is to decide deliverable formats and data management methods before starting work. This is often overlooked but is extremely important in cultural heritage 3D data creation. Data production itself is not the goal; the meaning lies in preserving, using, and handing the data on. Even if acquisition and editing go well, if the delivered files cannot be opened, are too heavy to handle, leave unclear which is the latest version, or lack spatial reference, the practical value drops dramatically.

3D data for cultural heritage consists of multiple layers: raw source data, intermediate edited data, usable data for applications, and lightweight viewing data. Delivering everything as an undifferentiated bundle makes it hard for users to handle. Separating what should be preserved, what is convenient for daily use, and what should be optimized for public viewing already greatly improves later usability.

File naming and folder structure rules are also important. If you do not organize object names, acquisition dates, version numbers, processing stage, presence of coordinates, and usage categories, it becomes difficult to reuse data when personnel changes. Because cultural heritage is often intended for long-term preservation, use a structure that remains meaningful years later rather than only being understandable at the moment. Linking reports, photos, location maps, field notes, and processing conditions with the data also greatly increases future value.

Furthermore, consider the deliverable formats in relation to intended uses. For viewing-focused applications, ease of handling matters, while for research or comparative purposes preserving preprocessed data is essential. Delivering only lightweight public-facing data may be insufficient for future detailed analysis, while delivering only raw data may hinder everyday sharing and viewing. For cultural heritage projects, a two- or three-tier delivery design that addresses both needs is realistic.

Decide management methods in advance as well: storage location, backups, access rights, and how to record update histories. If you deliver without considering data size and viewing environment constraints, the recipient may be unable to store the data. Cultural heritage datasets are not one-off outputs but assets for the future. Therefore, incorporate deliverable formats and management methods from the start of project design, not as an afterthought.

Precaution 6: Plan for ongoing operation and updates

The sixth precaution is to avoid treating the project as a one-off and instead plan for ongoing operation and update mechanisms. Creating 3D data for cultural heritage is not a one-time job; its value increases over time. There is potential to broaden future use—comparing pre- and post-repair conditions, tracking changes over time, integrating with additional investigations, updating exhibitions, educational use, and organizing public archives. Therefore, it is important to consider from the initial acquisition how the data will connect to subsequent work.

If you overlook this, the current project may succeed while future comparisons become impossible. Causes include ambiguous acquisition ranges, unclear reference positions, missing records of processing conditions, or changes in data formats that reduce value as a continuous record. In cultural heritage practice, the ability to compare over time can be more important than immediate appearance.

To plan for ongoing operation, leave operational rules that make re-acquisition easy. Document which areas were targeted, which parts were prioritized, site conditions at the time, the criteria used for alignment, and what processing was applied. Such records make it easier for the next team to work under the same assumptions. As these records accumulate, the data evolves from mere 3D files into operational assets for cultural heritage management.

Balancing viewing and field use is also important in ongoing operation. Even high-quality data will not permeate daily work if it is hard to access on site; conversely, a convenient viewing environment alone cannot support detailed decision-making. Practitioners should plan realistic operations anticipating multiple usage stages—preservation, research, and sharing—so that the system is sustainable.

Success in cultural heritage 3D data creation is not simply producing good data. It is about making data that leads to subsequent decisions and records. Projects designed for ongoing operation have far higher practical value than one-off jobs.

How to proceed to avoid failure in cultural heritage 3D data creation

We have covered six precautions, but in practice the order in which you organize them matters. Standardizing the workflow helps keep decisions consistent and reduces uncertainty.

The very first step is to clarify the purpose. Determine whether the project is for preservation, comparison, or use, and share among stakeholders what will constitute success. If you can confirm users, use cases, intended duration, and whether the data will be public at this stage, you will reduce confusion downstream. Next, specify required accuracy and acquisition scope concretely. Decide how much of the object to capture overall, which parts require high-density recording, and whether you need to make the data comparable later.

Then confirm site conditions. Understand working hours, access paths, restrictions on areas of entry, surrounding obstacles, weather impacts, and lighting conditions, and create an acquisition plan that is feasible. Identify places that are easy to capture as well as areas prone to omission. In the field, separate overall acquisition and focused acquisition, and always allocate time to check for insufficiencies. Fieldwork should include not only capturing but also verifying and supplementing data where needed.

After fieldwork, divide processing steps according to purpose. Organize source data prioritized for preservation, base data for comparison and analysis, and viewing-friendly data for sharing. Record what processing was performed and what was omitted to facilitate future verification and reuse. Also, prepare file structures, naming conventions, version control, storage locations, and backup policies to reduce post-delivery confusion.

The key is to design the workflow as a connected sequence. If purpose definition and delivery design are disconnected, it becomes unclear what to capture. If fieldwork and post-processing are disconnected, deficiencies will surface during editing. If delivery and operation are disconnected, the data will not be used. Creating 3D data for cultural heritage should be planned as a flow, not as isolated points.

It is not necessary for practitioners to make all technical judgments themselves. The important part is not missing the decision points. If you can answer why you are creating the data, how much is needed, which conditions will constrain you, and in what form it should be preserved, you will largely be able to determine technology and team structures.

Recently, demand has increased for efficient on-site recording while maintaining positional awareness. To link photos, point clouds, drawings, and registry information while capturing the surrounding environment and object location, handling position information cannot be ignored. Especially for wide-area surveys or records across multiple sites, on-site position awareness and manageable data organization greatly affect downstream efficiency.

In this regard, high-precision positioning systems that are easy to use on site pair well with cultural heritage practice. For example, an iPhone-mounted GNSS high-precision positioning device like LRTK can quickly obtain position information on site and link recorded objects and related materials to their locations, improving organization. Such portable solutions can enhance recording accuracy and manageability without requiring large-scale systems, making them accessible options for practitioners seeking to streamline survey support and surrounding records. If you want to improve the overall quality of field records rather than treating 3D data creation as a standalone task, considering position information in the recording infrastructure will raise the overall project quality.

Conclusion

To avoid failure in creating 3D data for cultural heritage, resist being distracted by new or flashy technologies and instead carefully build the necessary practical design. Define the purpose, concretize required accuracy, understand site conditions, plan acquisitions assuming missing data and noise, decide deliverable formats and management methods in advance, and plan for ongoing operation. Grasping these six points greatly increases the chance of project success.

No two cultural heritage objects are identical, and there is no universally correct recording method. However, common pitfalls exist. That is why it is important to master project design basics before diving into specific technical debates. A single on-site decision can greatly influence future preservation, investigation, and reuse. 3D data has meaning only when it is preserved and left in a usable form for the future.

If you are about to undertake 3D data creation for cultural heritage, start by organizing the purpose and operation for your project. If you want to make field records more accurate and efficient, consider not only the 3D data itself but also establishing a recording foundation that includes position information. To raise recording reproducibility and operational usability for cultural heritage surveys and surrounding records, combining portable high-precision positioning devices like LRTK with 3D data workflows can further enhance outcomes. As a first step to reliably connect cultural heritage to the future, build a practical system that balances 3D data creation with on-site positioning and a sustainable workflow.