Drone Heritage Site Inspection and Documentation

Geocartis carries out aerial documentation and condition recording of heritage structures and sites in India, and survey of the areas around protected monuments for regulatory purposes. We produce measured 3D records, rectified elevations, roof and upper-level condition surveys, and dated baselines for conservation work, all from outside the structure with nothing touching it. We are based in Ahmedabad, a World Heritage City, and work nationally.

We are a survey and documentation services company. We do not build drones, we are not conservation architects, and we do not carry out conservation work.

Read this first: two permissions, and one of them is not ours to get

Heritage work in India needs two separate clearances and neither is automatic.

DGCA airspace clearance. Green zone allows operation to 120 metres above ground level without prior flight permission, yellow zone requires permission and carries a lower ceiling, red zone is closed. A great many of India’s best-known monuments sit inside dense cities, near airports, or within areas that are restricted for other reasons. Some of the most famous cannot be flown by any commercial operator.

Permission from the monument’s custodian. For a centrally protected monument that is the Archaeological Survey of India, which maintains roughly 3,700 monuments of national importance. For a state-protected monument it is the state archaeology department. For a private property, a trust, a temple or a church, it is the owner. Aerial photography at protected monuments is controlled by the custodian and no DGCA clearance substitutes for their permission.

We do not obtain ASI permission on a client’s behalf and we would be sceptical of anybody who claims they can. Where a project genuinely requires aerial documentation of a centrally protected monument, the permission is sought by the commissioning body, usually an institution, a conservation consultant working under ASI, or ASI itself. Our role is to fly once that is in place, to the conditions attached to it.

We say this at the top because the commonest heritage enquiry we receive is from somebody who wants aerial imagery of a famous monument and has not considered either permission. It is better to establish in the first conversation that the answer may be no.

Where the work is more straightforward: state-protected monuments where the department is the client, unprotected and locally listed heritage, private palaces and havelis, religious trusts that own their own property, municipal heritage precincts, and heritage buildings in institutional ownership. That is a very large body of built heritage in India and most of it has never been properly documented.

The other market, and it is probably the bigger one

Here is a use of aerial survey around monuments that requires no permission to photograph the monument at all, and it is the one most likely to pay.

The AMASR Act 1958, as amended in 2010, creates a prohibited area extending 100 metres in all directions from the boundary of a centrally protected monument, in which construction is barred, and a regulated area extending a further 200 metres beyond that, in which construction requires permission from the National Monuments Authority or the competent authority. The 2017 amendment allowed public works within the prohibited area under defined conditions. The NMA, constituted under the 2010 amendment, grants those permissions and frames heritage bye-laws for individual monuments.

That framework creates a measurement problem that somebody has to solve on every application. Where exactly is the monument boundary. How far is the proposed construction from it. Which parts of a site fall inside 100 metres, which fall inside 300, and which are clear. What is already built there and when.

An accurate, dated, georeferenced survey of the area around a monument answers all of that, and the flight happens over the surroundings rather than over the monument. Distances are measured rather than scaled off a drawing, the existing built condition is documented, and the whole thing is presented in a form an NOC application or an objection to one can rely on.

Who needs this: developers and their architects applying for NMA clearance, urban local bodies planning works near a protected monument, infrastructure agencies routing a road, metro or utility past one, and ASI or NMA themselves when assessing encroachment.

It is also useful the other way round. Encroachment into a prohibited area is a recurring problem, and a dated survey is the evidence base for doing something about it.

We are careful about what this establishes. We measure distance from the boundary the custodian identifies to us. We do not determine where the protected boundary legally runs, and where the boundary is itself disputed or unmarked on the ground, we say so and the report records what we measured from.

What aerial documentation of a structure produces

Measured 3D record. A photogrammetric point cloud and textured mesh of the structure’s exterior, from which plans, elevations and sections can be derived. For many Indian heritage buildings no measured drawings exist at all, and where they do they are often a century old and were made by hand.

Rectified elevations and orthophotos. A geometrically corrected image of a facade, to scale, on which condition can be mapped directly. Conservation architects work on these, and producing them from aerial capture is far quicker than from scaffolding.

Roof, dome, spire and upper-level condition. The parts nobody can see. Roof tiles and terracing, dome and shikhara surfaces, finials, parapets, chhatris, drainage outlets and gargoyles, upper carving, vegetation growth in masonry joints, and water pooling. On most heritage structures the upper surfaces are where decay begins, because that is where water gets in, and they are the least inspected part of the building.

Vegetation and biological growth mapping. Peepal and other self-seeded growth in masonry is one of the most destructive agents on Indian heritage structures, and it establishes itself where nobody looks. Aerial survey finds it early.

Structural deformation baseline. A dated 3D record against which a later capture can be compared, showing movement, leaning, bulging or settlement over years. This only works if both epochs sit on the same survey control, which is why the first capture needs to be set up properly even though its value is not realised for a decade.

Site and landscape context. The monument within its setting, including approaches, boundary walls, water bodies, tanks, gardens, subsidiary structures and the encroachment around it.

Archaeological site survey. Mound topography, surface features, excavation trench recording, site plans and the extent of a site that is not obvious at ground level.

Pre- and post-conservation record. Documentation before an intervention and after it, which is basic conservation practice and frequently not done to a measurable standard.

Post-disaster documentation. After an earthquake, fire, flood or cyclone, rapid capture of a damaged structure before anything is moved. This is time-critical and it is one of the strongest arguments for a heritage body having a relationship with a survey provider in advance rather than looking for one afterwards.

Risk to the fabric, which we take seriously

This deserves its own section because it separates people who should do this work from people who should not.

A drone that comes down on a construction site damages a drone. A drone that comes down on a five hundred year old carved stone facade damages something irreplaceable, and no insurance policy restores it.

So heritage flying is done differently from everything else we do. Larger standoff distances than the imagery strictly requires, with detail obtained through optics rather than proximity. Conservative wind limits, lower than for any other work, because turbulence around a large masonry structure in a built-up area is unpredictable. No flight directly over a fragile element where a failure would drop the aircraft onto it. Approach and departure paths planned so that a loss of control at any point takes the aircraft away from the structure rather than into it.

And a willingness to abort. On heritage work we would rather return another day than take a marginal pass. That costs a mobilisation. The alternative cost is not measurable in money.

Living sites need more than technical care. Many Indian heritage structures are active places of worship with worshippers present, and many are visitor attractions with crowds. A drone overhead is intrusive, and in some contexts it is offensive. We work through the custodian on timing, we avoid prayer times and festivals, and we accept a restriction on what may be photographed where one is given. Where a site’s community objects, that ends the matter as far as we are concerned, whatever permission exists on paper.

How we do it

Step 1: Establish what is possible before anything else

The first exercise on a heritage enquiry is not a quotation, it is a feasibility check: the airspace classification of the site, the protection status of the structure, who the custodian is, and whether permission for aerial work is realistically obtainable.

We will tell you if the answer looks like no. On centrally protected monuments in restricted airspace it often is.

Step 2: Documentation brief

Heritage documentation is only useful if it answers a defined question, and the question determines the method.

A record for a conservation management plan, a base for measured drawings, a condition survey of the roof, a deformation baseline, a pre-intervention record, a virtual access model and an NOC distance survey are six different jobs with different capture requirements and different resolutions. We settle which before planning.

Where a conservation architect or archaeologist is leading, we work to their brief and their conventions. Documentation produced in a form that suits us rather than the discipline is documentation somebody has to redo.

Step 3: Control and reference

Control points are established on stable ground around the structure, sited so the block is constrained on all sides. On a tall structure in a tight urban setting, that can be difficult and it is the part of the job that takes the time.

Nothing is fixed to the structure. Control markers are placed on the ground, on modern paving, or on adjacent modern surfaces, never adhered, pinned or attached to historic fabric, and nothing is left behind.

Where a long-term deformation baseline is the purpose, permanent control is established off the structure in positions that can be reoccupied in ten or twenty years, documented with photographs and coordinates in a form that survives staff turnover at the institution.

Independent checkpoints are withheld from the adjustment and the measured RMSE is stated.

Step 4: Flight planning

Ground sample distance is pixel pitch multiplied by flying height divided by focal length. A 20 megapixel one-inch sensor with 5472 pixels across 13.2 millimetres has a pixel pitch of about 2.41 micrometres, and with an 8.8 millimetre lens at 120 metres gives roughly 3.3 centimetres GSD for site-level work.

Structural documentation needs far finer than that, and it is achieved by flying closer and by optics rather than by dropping altitude carelessly near a building.

Heritage capture is overwhelmingly oblique. A nadir block gives roofs. Everything that makes a heritage structure what it is, the facades, the carving, the jharokhas, the columns, the shikhara faces, is vertical, and it is captured with oblique passes at multiple headings and multiple elevations. On a complex structure that means a systematic orbit pattern at several heights rather than a single pass.

Overlap runs high, because ornate carved surfaces with deep shadow are harder for tie point matching than they look, and because gaps in a heritage record are expensive to fill later.

Step 5: Capture

Light is a genuine constraint on carved stone. Hard direct sun produces deep shadow in carving and blown highlights on light stone, and both hide detail. Overcast or soft light is better for facade documentation, and on east or west facing elevations the timing window is narrow.

Visitors and worshippers determine when we can work. Early morning before opening is often the only viable window at an active site, and that has to be arranged with the custodian.

Birds are a real hazard at heritage sites more than anywhere else. Pigeons, parakeets and raptors nest in and around old structures in numbers and they respond to a drone.

The crew works without touching anything, without placing equipment against the structure, and without leaving any trace.

Step 6: Processing

Aerial triangulation, dense matching, mesh generation and texturing. For heritage the mesh matters more than it does on other work, because the deliverable is often a visual and geometric record rather than a measurement of quantities.

Rectified elevations are generated by projecting the model onto defined planes, one per facade, at a stated scale, and those become the base drawings a conservation architect annotates.

Where derived drawings are in scope, plans, elevations and sections are cut from the point cloud. Whether those become finished measured drawings depends on the level of detail agreed, and a drawing derived from photogrammetry captures what the sensor saw rather than what an architect would draw. We are clear about which we are supplying.

Step 7: Reporting

The record, the derived products, the control and accuracy statement, the capture conditions, and a statement of what was and was not covered. That last point matters on heritage: the parts of a structure that could not be seen, because of adjacent buildings, tree cover, restricted airspace or safe standoff, are listed rather than quietly interpolated.

Accuracy, and where a drone is the wrong tool

For a well-controlled photogrammetric block, horizontal RMSE typically lands between one and two times the GSD and vertical between two and three times, and we report the measured figures from withheld checkpoints.

Terrestrial laser scanning is the standard for high-precision heritage documentation and we will say so when it is what you need. A scanner set up around and inside a structure produces denser, more consistent geometry than aerial photogrammetry, particularly on interiors, on complex intersecting geometry, and where millimetre-level precision is required for structural analysis or for replicating a component.

The two are complementary rather than competing. A scanner cannot see a dome from above, cannot reach a spire, and cannot cover a large site efficiently. A drone cannot go inside, cannot see under a canopy or a chhajja, and cannot match a scanner’s precision on a detailed element.

The sensible arrangement on a serious documentation project is both, registered to common control, with aerial capture covering roofs, upper levels and site context and terrestrial scanning covering interiors and detail. Where a client is buying only one, we will tell them honestly which one their brief actually needs, including when the answer is not us.

What aerial photogrammetry does not do on heritage: interiors, anything under a projecting element, subsurface archaeology, material characterisation, structural assessment, or documentation of a surface that cannot be seen from the air.

Deliverables and output file formats

Raw imagery is handed over. On heritage work the archive has value beyond the immediate project, and it should sit with the custodian rather than with us.

Categories of client, and what each needs

Centrally protected monuments. ASI is the custodian and the permission route runs through them. Realistically we work here as a subcontractor to an institution or consultant already engaged, not as a direct approach.

State-protected monuments. State archaeology departments hold a large number of monuments with far less documentation than the ASI estate and far less capacity to produce it. This is more accessible work.

Listed and unprotected heritage. Locally listed buildings, precincts under municipal heritage regulations, and the enormous body of significant built heritage with no formal protection at all. Municipal heritage cells, INTACH chapters and conservation practices work in this space.

Private heritage. Palaces, havelis, forts and country houses, many now hotels or in family ownership, where documentation supports restoration, insurance, valuation or adaptive reuse.

Religious trusts. Temples, mosques, churches, gurdwaras, jain derasars and dargahs that own and maintain their own structures. Substantial ongoing conservation happens here, largely undocumented.

Educational and research institutions. Universities, archaeology departments and research projects.

Developers and authorities near monuments. The NOC distance survey described above.

Where we work

We are based in Ahmedabad, which is a reasonable place to run a heritage practice from. Gujarat holds four UNESCO World Heritage properties: Champaner-Pavagadh Archaeological Park, Rani ki Vav at Patan, the Historic City of Ahmedabad, which was India’s first World Heritage City, and Dholavira. India as a whole now has forty-five World Heritage properties following the inscription of Sarnath in July 2026.

Gujarat. The walled city of Ahmedabad and its pols, Champaner and Pavagadh, Patan, Dholavira and the Kutch sites, Modhera, Siddhpur, Junagadh, Vadnagar, Lothal, Sarkhej Roza, and the palace and haveli heritage across Saurashtra and north Gujarat.

Rajasthan. The fort and palace heritage of Jaipur, Jodhpur, Udaipur, Jaisalmer, Bundi, Chittorgarh and Shekhawati, and a very large body of privately held heritage property.

Northern India. Delhi, Agra and the Braj region, Lucknow, Varanasi, Amritsar, Punjab and the hill states.

Central and western India. Madhya Pradesh including Khajuraho, Orchha, Mandu, Gwalior and Sanchi, and Maharashtra including Mumbai’s precincts, the Deccan forts, and the cave sites.

Southern India. Temple architecture across Tamil Nadu, Karnataka including Hampi, Badami and Belur, Kerala, Telangana and Andhra Pradesh.

Eastern India. Bengal terracotta temples, Odisha’s temple architecture, and the heritage of Kolkata, Murshidabad and the north-east.

A note on protection status: it varies within a single property. The Maratha Military Landscapes inscribed in 2025 include forts under ASI protection and forts under Maharashtra state archaeology, so the permission route differs across one World Heritage property. Establishing who the custodian is for each element comes before anything else.

What we do not do

We do not obtain ASI permission, and we do not fly at a protected monument without the custodian’s permission being in place.

We do not document interiors.

We do not carry out structural assessment, and we do not advise on structural stability.

We do not identify materials, analyse mortar, characterise stone decay mechanisms or diagnose biological agents. Those require sampling and a conservation scientist.

We do not prepare conservation plans, conservation management plans or restoration proposals.

We do not carry out subsurface archaeological investigation. Ground penetrating radar and magnetometry are separate techniques.

We do not attach, fix, adhere or lean anything on historic fabric.

We do not fly closer to a structure than a safe standoff, and we abort rather than take a marginal pass.

We do not fly over an active place of worship against the wishes of its community.

We do not fly in red zones.

Frequently asked questions

Can you fly over a famous monument? Often not. It needs both DGCA airspace clearance and the custodian’s permission, and many well-known monuments sit in restricted airspace or have aerial photography controlled by ASI. We will check and tell you honestly, usually in the first conversation.

Can you get us ASI permission? No. That is sought by the commissioning institution or consultant. We fly once it is in place and to whatever conditions it carries.

Do we need permission to survey around a monument for an NOC application? DGCA airspace clearance applies as anywhere. The AMASR prohibited and regulated area provisions govern construction, not flying, so a survey of the surroundings is a different proposition from photographing the monument. We would still inform the local ASI office as a matter of courtesy and good practice.

Can this establish how far our site is from a protected monument? Yes, measured rather than scaled from a drawing, from the boundary the custodian identifies. Where the protected boundary is disputed or not marked on the ground, we record what we measured from and say so. We do not determine where the legal boundary runs.

Is a drone safe near a fragile structure? Only with deliberate procedure: larger standoff than the imagery needs, optics rather than proximity, conservative wind limits, planned failure paths away from the structure, and a willingness to abort. On heritage we would rather lose a day than risk a facade.

Can you produce measured drawings? We can produce rectified elevations and cut plans, elevations and sections from the point cloud, to an agreed level of detail. Whether that constitutes finished measured drawings depends on the brief. Photogrammetry captures what the sensor saw, which is not the same as what a conservation architect would draw, and we are clear about which we are supplying.

Should we use laser scanning instead? For interiors, for millimetre-precision detail and for complex intersecting geometry, yes. For roofs, domes, spires, upper facades and site context, aerial photogrammetry reaches what a scanner cannot. On a serious documentation project both are used, registered to common control, and we will tell you if your brief actually needs the scanner rather than us.

Can you record a structure before conservation work starts? Yes, and it is one of the most useful things you can commission. A dated measured record before intervention is basic conservation practice and it is regularly skipped for cost reasons that look small next to the value of having it.

Can you document damage after an earthquake or a fire? Yes, and quickly, which matters because the record needs to exist before anything is moved or shored. A heritage body that has this arranged before an event is in a much better position than one arranging it during.

What file formats do you deliver? Mesh as OBJ or FBX. Point cloud as LAS, LAZ and E57, which registers with terrestrial scan data. Rectified elevations as TIFF, PDF and DWG. Drawings as DWG and DXF. Orthomosaic and terrain as GeoTIFF. Reports as PDF.

Get in touch about your site

Tell us the structure or site, its protection status, who the custodian is, and what the documentation is for: a conservation plan, measured drawings, a condition survey, a deformation baseline, a pre-intervention record, or an NOC distance survey.

We will check the airspace position and the permission route and tell you what is feasible. If the answer is that the site cannot be flown, you will hear it from us early rather than after you have committed to a programme.