Drone Bridge Construction Progress Inspection

Geocartis monitors bridge construction across India: foundation and pier progress, superstructure erection span by span, segment casting yard output, approach earthwork and quantities, and a dated georeferenced record of the whole structure at each capture. Findings are reported against your own pier and span numbering. We are based in Ahmedabad and work nationally.

We are a survey and inspection services company. We do not build drones, we do not sell software, and we do not carry out structural assessment.

Why bridges are harder to monitor than anything else on a project

Every large civil project has progress reporting problems. Bridges have a specific one: there is nowhere to stand.

On a building site you can walk the floors. On a road package you can drive the alignment. On a bridge, the work is happening on top of a pier 40 metres above a river, or on a deck that has no access from the ground, or on a segment being erected from a launching girder that nobody can approach. The project manager’s picture of progress comes from what the site engineer reports, and the site engineer’s picture comes from what they can reach.

So bridge progress reporting tends to be verbal, optimistic and unverifiable. A pier is reported as complete when it is nearly complete. A span is reported as erected when the segments are placed but the stressing is not done. Nobody is lying; there is simply no independent picture.

A drone gives you the vantage point that does not otherwise exist. Deck level from above, pier shafts from the side, the casting yard, the approaches and the river crossing, all in one capture, on a stated date.

The commercial case: milestones and chronology

Two things pay for this service.

Milestone verification. Bridge packages are commonly paid against physical milestones: foundations complete, piers cast to a stage, spans erected, deck cast. Those are countable, and counting them from the air is quick and unambiguous. For an employer, a lender’s engineer, or a PMC monitoring a package remotely, an independent monthly count against the claimed milestone is a straightforward control.

Chronology, which matters more than anybody expects at the start. River bridges in India are delayed. Flood seasons stop work, high water prevents access to foundations, and a season lost is a year lost on a river crossing. Extension of time claims on bridge packages routinely turn on when a work front stopped, when the water came up, when it went down, and when work restarted.

A monthly dated georeferenced record answers all of that. It is the single most valuable thing a bridge contractor or an employer can have when the programme is being argued three years later, and it costs a fraction of what the argument costs.

On river bridges specifically, we push for a capture immediately before the monsoon and another immediately after. Those two datasets bracket the flood season and establish exactly what state the works were in going into it and coming out.

What we capture at each stage

Site before works start. Original ground level across the site and approaches, the river channel and banks at a stated water level, existing structures, adjacent property, access and the condition of anything that a neighbour might later claim was damaged. This is the baseline everything else measures against, and on a bridge it is also the record of what the river looked like before the piers went in.

Foundations. Pile positions and pile cap progress where above water, open foundation excavation, well foundation curb and steining above water level, cofferdam construction and condition, and the working platforms and temporary works around them. What is below water is not visible, and that limit is set out below.

Substructure. Pier and abutment progress by lift, formwork position, pier cap construction, reinforcement cages at the casting stage, bearing pedestal work, and pier geometry as built.

Superstructure. Girder or segment erection span by span, launching girder position and progress along the alignment, cantilever construction at each pier, deck slab casting, diaphragm and cross girder work, and stressing operations where visible.

Deck and finishing. Wearing coat progress, expansion joint positions, crash barrier and railing, drainage spouts, lighting and services, and approach slab work.

Approaches. Embankment earthwork with cut and fill quantities by chainage, retaining and reinforced earth wall progress, subgrade and pavement layers, and the transition to the main carriageway.

Casting yard. This one is under-used and it is the neatest application on the whole page. A precast segmental bridge is fed by a casting yard, and the yard’s output is countable from the air: segments cast, segments stored by type, stockyard occupancy and the rate of production over successive captures. A yard falling behind shows up in the segment count weeks before it shows up as an erection delay, which makes it a leading indicator rather than a lagging one.

Temporary works and site logistics. Staging, cranes, launching gantry, access roads, batching plant, material stacking and river diversions.

Construction methods and what each one needs

The capture plan follows the erection method.

Precast segmental with launching girder. Track the girder’s position along the alignment, span completion, and segment counts in the yard. Progress is naturally measured in spans and the aerial record maps directly onto it.

Balanced cantilever. Cantilever arm progress from each pier, symmetry of the two arms, form traveller position, and closure pour status. Obliques are essential because the interesting work is on the underside and the ends of the cantilever.

Incremental launching. Casting bed activity behind the abutment, launch nose position, and the position of the deck along the alignment at each capture.

Cast in situ on staging. Staging erection, formwork, reinforcement and pour sequence. The staging itself is worth recording, because it is temporary works that gets dismantled and later disputed.

Steel girder launching. Girder assembly on the bank, launching progress, splice completion and deck slab work over the steel.

Cable-stayed and extradosed. Pylon construction by lift, stay cable installation sequence, deck segment progress from the pylon outward. Flying near a cable-stayed structure needs particular care and that is covered below.

Geometry, as-built and where the limits are

We can report deck level and pier position as built, compared against design, and that comparison is useful for spotting a systematic problem across several spans.

It is not a substitute for setting-out survey or precise levelling. Photogrammetry at 5 cm ground sample distance resolves to roughly 5 to 10 centimetres horizontally and 10 to 15 centimetres vertically. Bridge construction tolerances on pier position, bearing level and deck profile are in millimetres, controlled with total stations and precise levels. Drone data is two orders of magnitude coarser and should never be presented as capable of that.

What it does do is find the errors big enough to matter across a length: a run of spans deviating from the design profile, an approach embankment built systematically low, an alignment drift over several piers. Those are the failures that cost money and that spot-checking misses.

And it does not measure deflection. Camber, deflection under load, and long-term creep and shrinkage movement are monitoring exercises with instruments, not aerial photogrammetry.

What a drone cannot do on a bridge

Nothing below the waterline. Pier foundations under water, well foundation sinking below water level, scour around piers, riverbed condition and any underwater element are outside what a camera reaches. Scour assessment in particular needs bathymetric survey or a diver inspection, and on a river bridge scour is one of the things that most matters. We say what we measured and where the measurement stops.

Limited soffit access. The underside of a deck is in GNSS shadow, because the deck itself blocks the sky. Flying under a deck is a GNSS-denied operation, and it is not something an ordinary survey platform does safely. Where soffit condition is required, that is close-range work with a suitable aircraft and a pilot experienced in it, or it is access equipment. We will tell you which rather than attempting it.

No non-destructive testing. Rebound hammer, ultrasonic pulse velocity, half-cell potential, cover meter, core sampling and chloride testing are all contact methods on the structure.

No structural or condition assessment. We record and measure. Assessment of a structure’s condition, capacity or fitness is a bridge engineer’s work.

No load rating and no load testing. Load testing follows its own IRC guidance and is an instrumented exercise.

Nothing internal. Box girder interiors, duct condition, grouting quality, bearing internals and expansion joint internals are all inaccessible.

In-service bridges: where the boundary sits

Worth addressing, because clients often assume progress inspection and condition inspection are the same service with a different subject.

Inspection and maintenance of bridges in India is framed by IRC guidance, principally IRC:SP:35 on inspection and maintenance, with IRC:SP:40 covering strengthening and rehabilitation and setting out condition states from excellent through good, fair and poor to critical, and IRC:SP:51 covering load testing. On National Highways, MoRTH’s Indian Bridge Management System provides the inventory and condition rating framework. The sequence in that guidance runs from routine visual inspection, through condition evaluation, to detailed investigation and special testing where the evaluation warrants it.

A drone contributes to the first step and not to the rest. High resolution imagery of surfaces that a person cannot reach, the tops of piers, the outer faces, the soffit where access allows, the bearings from above, is genuinely useful visual inspection input, and it is safer and faster than the alternatives.

But the condition rating is an engineering judgement made by a qualified inspection engineer against the guidance, informed by testing that a camera cannot perform. We can provide the imagery that engineer works from. We do not produce the rating, and any provider offering to deliver a condition rating from aerial imagery alone is offering something the guidance does not support.

Where a client wants both, the sensible arrangement is that we capture and their bridge engineer or inspection consultant assesses. We are comfortable working under an inspection engineer’s direction on exactly that basis.

How we do it

Step 1: Structure data and the reporting frame

We ask for the general arrangement drawing, the pier and span numbering, the chainage, the construction methodology and erection sequence, the programme, the milestone definitions in the contract, and any previous survey.

Every deliverable is reported against your pier and span numbering. A report that refers to “the fourth pier from the left bank” is a report somebody has to translate.

Where an as-built comparison is in scope, we need the design surface or setting-out data and the survey control the project is using, so the transformation between the project grid and our survey frame can be established properly. That is discussed in more detail on our construction survey page and it applies identically here.

Step 2: Airspace, and the height question

We check the site against the Digital Sky airspace map before quoting. Green zone to 120 metres above ground level without prior flight permission, yellow zone with permission and a reduced ceiling, red zone not at all.

Two bridge-specific complications.

Tall structures. A high pier, a pylon on a cable-stayed bridge, or a crossing over a deep valley can put the top of the structure near or above the green zone ceiling measured from the ground below. That needs permission and it needs establishing before mobilisation.

Boundaries and sensitivity. River bridges frequently form or cross state and district boundaries, which multiplies the administrative coordination. Bridges in border regions and near defence installations carry restrictions that no commercial operator can work around.

Our aircraft carry active Unique Identification Numbers and our pilots hold DGCA Remote Pilot Certificates.

Step 3: Site coordination

Bridge sites are congested, vertical and busy. We work to the project’s safety induction and permit system, and we coordinate with the site on crane movements, launching operations, stressing operations and any live traffic on an adjacent or existing structure.

Where the bridge carries or crosses live traffic or a live railway, the relevant authority’s permission is needed in addition to the project’s, and the railway case in particular has its own lead time.

Step 4: Control

Control goes on stable ground on both banks and, where the structure is long, at intermediate points on completed works or on the approaches. Bank-only control on a long river crossing constrains the ends and leaves the middle weakly held, which produces exactly the kind of smooth error that looks plausible and is not.

On a monitoring programme we establish permanent control that survives the whole project, sited clear of the works and the approach earthworks, and re-occupy it at every capture. Epoch-to-epoch comparison is only meaningful if both epochs sit on the same frame.

Independent checkpoints are withheld from the adjustment and the measured RMSE is reported at every epoch.

Step 5: 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 that gives roughly 3.3 centimetres GSD for the site-level capture.

Obliques are the core of bridge work, not an add-on. A bridge is a vertical structure and a nadir block records the deck and essentially nothing else. Pier shafts, pier caps, bearings, cantilever ends, abutments, wing walls and the outer faces of the deck all require oblique passes at multiple headings and heights. On a tall pier that means passes at several elevations.

Flight plans are repeated between epochs, same lines, same altitudes, same oblique positions, so that this month compares against last month rather than merely resembling it.

Cable-stayed structures need specific planning. Stay cables are thin, under tension, and close to invisible against a bright sky. They are a collision hazard of the same character as an overhead conductor and worse, because there are many of them in a fan geometry. On stayed and extradosed structures we plan approach and transit paths that do not cross the cable plane, and we treat the fan as a solid exclusion volume rather than something to fly between.

Step 6: Capture

Over water, assume anything lost is gone. A drone that comes down on the bank is recoverable; one that comes down in a river in spate is not. Flight lines over open water are planned to minimise exposure, battery reserves are set conservatively, and we do not push a marginal pass over the channel.

Wind behaves badly at bridge sites. River valleys, gorges and the gap between high approach embankments all funnel and accelerate wind, and the structure itself creates turbulence. Close-range oblique work near a pier in gusty conditions is where a bridge job goes wrong, and our limits for close work are lower than for open mapping.

Cranes and launching gantries are marked as exclusion volumes and their operators notified. A launching girder in motion is a large moving structure and flying near it during a launch is not something we do.

Water level is recorded at every capture, because it determines how much of the substructure was visible and it is what makes two epochs comparable.

Step 7: Processing and reporting

Aerial triangulation and dense matching for the measured outputs, oblique imagery indexed by structure element for the visual record.

The report is organised by pier and span. For each, the status at this capture, what changed since the last, and imagery. Quantities on approaches and earthwork are reported by chainage segment. Segment counts from the casting yard are reported as a running total against the requirement.

We separate what is measured from what is observed. A span whose segments are all placed is observable. Whether it is stressed and structurally complete usually is not, and the report says so rather than recording the span as complete.

On a monthly programme, the format stays identical between epochs, and the summary shows progress against the programme activities that are physically visible.

Deliverables and output file formats

Raw imagery is handed over with each epoch. On a structure where the programme may be argued years later, holding the raw data yourself matters.

Where we work

We are based in Ahmedabad and work nationally, structuring bridge work as a monitoring programme with scheduled mobilisation rather than single visits.

Gujarat. River crossings on the Narmada, Tapi, Mahi and Sabarmati, including the high speed rail river bridges and the Delhi-Mumbai Expressway structures, the coastal and Saurashtra crossings including the Okha area, Ahmedabad’s flyovers and Sabarmati bridges, and the metro viaducts across Ahmedabad and Gandhinagar.

Western India. Mumbai, Navi Mumbai, Thane, Pune, Nashik and Nagpur, including expressway structures and metro viaducts, plus the Konkan crossings and Goa.

Northern India. Delhi NCR, Uttar Pradesh, Punjab, Haryana, Rajasthan, Uttarakhand and Bihar, including Ganga, Yamuna, Ghaghara and Gandak crossings, via Lucknow, Kanpur, Prayagraj, Varanasi, Patna, Chandigarh and Dehradun.

Central India. Madhya Pradesh and Chhattisgarh, via Bhopal, Indore, Jabalpur and Raipur.

Eastern and north-eastern India. West Bengal, Odisha, Jharkhand and Assam, including the Hooghly, Mahanadi, Brahmani and Brahmaputra crossings, via Kolkata, Bhubaneswar, Ranchi and Guwahati.

Southern India. Telangana, Andhra Pradesh, Karnataka, Tamil Nadu and Kerala, via Hyderabad, Vijayawada, Visakhapatnam, Bengaluru, Chennai and Kochi.

What we do not do

We do not measure or assess anything below the waterline, including foundations, scour and riverbed condition.

We do not carry out non-destructive testing.

We do not produce condition ratings, structural assessments, capacity assessments or load ratings.

We do not carry out load testing.

We do not inspect box girder interiors, ducts, bearings internally, or expansion joint internals.

We do not measure deflection, camber under load, or long-term structural movement.

We do not carry out setting out or precise levelling, and our as-built comparison does not work to construction tolerance.

We do not fly under a deck as a routine survey operation, and we do not fly through or near a stay cable fan.

We do not fly above the green zone ceiling without permission, and we do not fly in red zones.

Frequently asked questions

Can you tell us how many spans are complete? We can tell you how many spans have their girders or segments placed, and record the state of each. Whether a span is structurally complete, stressed and accepted is a project determination, and the report distinguishes what is observable from what is not.

Can you check the piers are in the right position? We can compare as-built pier position against design and find deviations big enough to matter across several piers. We cannot verify pier position to construction tolerance, which is millimetre work with a total station.

Can you inspect the foundations? Only the parts above water and above ground. Everything below the waterline, including scour, needs bathymetric survey or divers.

Can you inspect the underside of the deck? Not as part of a standard survey. The soffit sits in GNSS shadow because the deck blocks the sky, which makes it a GNSS-denied close-range operation with a different aircraft and pilot skillset. We will tell you when the requirement needs that rather than attempting it.

Is this safe around a cable-stayed bridge? With specific planning. Stay cables are thin, numerous and hard to see against sky, and we treat the cable fan as a solid exclusion volume rather than flying between the cables. Nothing about the survey requires proximity to them.

Can you fly while the launching girder is operating? No. A launch is a large moving operation and we schedule around it, in coordination with the site.

How often should we fly? Monthly suits most bridge packages. On a river crossing we would add a capture immediately before the monsoon and another immediately after, because those two bracket the season that causes most bridge programme disputes.

Can this support an extension of time claim? It provides the dated, georeferenced factual record of what state the works were in and when. That is evidence into a claim, not a claim. It is also considerably better evidence than site photographs with no location, date verification or scale.

Can you inspect an existing bridge in service? We can capture high resolution imagery of surfaces that are hard to reach, which is useful input to a routine visual inspection under the IRC guidance. We do not produce condition ratings. The assessment is a qualified bridge inspection engineer’s, and we are happy to work under one.

What file formats do you deliver? Orthomosaic and surface models as GeoTIFF. Point cloud as LAS, LAZ and E57. Mesh as OBJ or FBX. Drawings and sections as DXF and DWG. Progress and quantity reports as PDF and XLSX. Imagery as JPEG, indexed by element.

Start a monitoring programme

Send us the general arrangement, the pier and span numbering, the construction methodology, the programme duration and the frequency you want. Tell us whether the crossing is over water, whether there is a casting yard to cover, and whether an adjacent structure carries live traffic or a railway.

We will return the airspace position including whether structure height requires permission above the green zone ceiling, the control plan, the epoch schedule, the reporting format, and a price for the programme rather than per visit.

If the structure has not started, book the baseline capture first. On a river crossing, book it before the monsoon.