Geocartis inspects overhead power lines across India for transmission utilities, distribution companies, generators and their contractors. We cover post-storm and post-outage patrol, routine tower and conductor condition survey, thermal hotspot detection at joints and clamps, right of way vegetation and encroachment mapping, and corridor survey for new lines. Nobody climbs a tower and nobody walks the line. We are based in Ahmedabad and work nationally.
We are a survey and geospatial services company. We do not build drones, we do not sell software, and we do not carry out live-line work of any kind.
What “finding the fault” actually means
Clients often describe this service as a drone that finds the fault on a line, and there is a version of that which is true and a version which is not. Getting the distinction right matters, because it changes what you should buy.
A drone does not locate an electrical fault. When a feeder trips, the location comes from your protection system: distance protection giving a distance-to-fault, fault passage indicators along the feeder, relay records, and the switching sequence your control room runs to narrow down the section. That is electrical engineering and a camera contributes nothing to it.
What a drone does is replace the patrol. Once the section is narrowed to a few kilometres, somebody has to go and look at it. In rural India that has traditionally meant a lineman on a motorcycle following a line across fields, canals, scrub and riverbeds, stopping at each pole or tower, and looking up. On a long feeder in difficult country that takes most of a day and it happens under time pressure with supply off.
A drone covers that section in a fraction of the time, sees the top of every structure rather than the underside, and produces a record. On the visible faults that cause most rural outages, this is genuinely decisive.
Faults a drone finds quickly and reliably: snapped or dropped conductor, fallen or leaning pole, broken cross arm, tree or branch on the line, flashed or shattered insulator discs, burnt or parted jumpers, bird nests bridging live parts, damaged or broken earth wire, storm debris across the span, and structural collapse.
Faults a drone will not find: anything internal to a cable, transformer, switchgear or joint, insulation breakdown with no external evidence, most underground faults, and intermittent faults that were not present at the time of the flight.
So the honest description of this service is faster patrol and better condition data, not electrical fault location. In practice that is worth more than what people usually ask for, because the largest value is not finding today’s fault faster. It is finding next year’s fault before it happens.
Storm patrol and emergency response
The application where the case is most obvious.
After a cyclone, a severe thunderstorm, a dust storm or flooding, a utility has multiple feeders out simultaneously, damage spread across a wide area, roads blocked, and a limited number of crews. The first job is not repair, it is knowing what is broken and where, so that crews and materials go to the right places in the right order.
A drone patrol answers that faster than a ground patrol can, and it answers it with evidence rather than a phone call. Within hours of conditions becoming flyable, you can have the damaged spans identified, the structures affected listed, the nature of the damage on each recorded with imagery, and the access condition for each location visible.
That last point is underrated. Knowing that a tower is down is useful. Knowing that the approach road to it is under water, and that the adjacent tower is intact, changes what you send and when.
For utilities in cyclone-exposed states, and for the Gujarat and Rajasthan lines that take dust storms, having this arranged in advance rather than organised during the emergency is the difference between it being useful and it being an idea somebody had afterwards. We are willing to hold a standing arrangement for this with utilities that want one.
Routine condition inspection: what each sensor finds
Preventive patrol is where the recurring value sits, and it is sensor-dependent.
High resolution RGB. Broken or damaged conductor strands, corrosion on tower members and hardware, cracked, chipped, flashed or contaminated insulator discs, missing or loose bolts and nuts, missing tower members, displaced or missing vibration dampers, spacer and spacer damper condition, arcing horn misalignment, jumper condition, bird nesting and bird guards, damaged earth wire and earthing connections, foundation condition and erosion around footings, tower tilt visible against the vertical, and damaged or missing danger plates and phase plates.
Thermal infrared. Hotspots at compression joints, mid-span joints, clamps, jumper connections, dead-end fittings and disconnector contacts. A resistive joint heats before it fails, so thermal patrol finds a class of defect that visual inspection cannot see at all. Thermal results need interpretation against load at the time of capture, ambient temperature, wind and emissivity, and a temperature rise measured on a lightly loaded feeder at 7 am means something different from the same reading at peak load. Any thermal report worth having states load and conditions.
LiDAR. Conductor position in three dimensions, ground clearance, clearance to vegetation, structures and crossings, and the geometry needed for sag and tension modelling. This is the sensor for clearance compliance and the section below explains why.
Corona and UV. Partial discharge on insulators and fittings, detected with a specialist daylight corona camera. It finds a different failure mode again, on contaminated or degraded insulation. It is a specialist payload and not part of a standard survey.
A note on what we currently fly is in the editorial section of this document. On the published page we state our sensors explicitly, because a client who asks for thermal and receives visual has been badly served.
The clearance problem, and why a measurement is not compliance
This is the technical point that separates a serious provider from a vendor selling flights, and most Indian clients have never had it explained.
Statutory minimum clearances for overhead lines are set by the CEA safety regulations by voltage, above ground and to structures and crossings. Your engineering department will hold the applicable figures for each line, and we work to those rather than to a number we have quoted from somewhere.
Here is the difficulty. A LiDAR survey measures where the conductor is at the moment of the flight. The conductor’s position depends on its temperature, which depends on the current it is carrying and on ambient conditions. A line surveyed at 10 am in December on light load sits considerably higher than the same line at 3 pm in May at maximum operating temperature. Wind adds lateral blow-out on top of that.
So a clearance measured on the day is not the clearance that matters. What matters is clearance at maximum operating temperature and, for lateral encroachment, under design wind. A survey that reports measured clearances against statutory minimums and calls it a compliance check will pass spans that are non-compliant under the conditions where compliance actually counts.
The correct method is to use the surveyed conductor positions to fit a sag-tension model for each span, using the conductor type, tension, span geometry and the temperature at survey, and then predict conductor position at the design maximum temperature and under design wind. Clearance is assessed against those predicted positions. Industry practice runs this through sag and tension software, and the deliverable is a span-by-span clearance assessment at design condition rather than a list of measurements.
We are explicit about which of the two you are buying. Where the scope is measured clearance on the day, that is what the report says. Where it is compliance assessment, the sag-tension modelling is in scope, we need the conductor and tension data from you, and the report states the design conditions modelled.
Any quotation you receive from anybody that promises clearance compliance without asking you for conductor type and stringing data cannot be doing the modelling.
Vegetation and right of way
The commonest cause of avoidable outages on wooded corridors, and the one that recurs every growing season.
We map vegetation within and adjacent to the right of way, with the position and height of encroaching growth, and where LiDAR is flown, the actual clearance from conductor to canopy. Output is a prioritised list by span rather than a general statement that the corridor needs cutting, so a tree-cutting crew is directed to the twenty spans that matter instead of walking the whole line.
Repeat capture across seasons shows growth rate, which is what lets a utility plan a cutting cycle rather than react to a trip.
Encroachment of a different kind matters too. Buildings, sheds, tube wells, hoardings, stacked material and construction under a line are a safety issue as much as a reliability one, and they are documented dated and positioned so an enforcement notice has evidence attached.
Transmission and distribution are different jobs
Worth separating, because the economics and the method differ.
Transmission and sub-transmission, 66 kV and above on towers. Long spans, defined right of way, high consequence of failure, and high value per kilometre. This is where detailed inspection, thermal, LiDAR clearance work and annual patrol cycles are justified, and where most drone inspection worldwide happens. Structures are tall, hardware is high, and a foot patrol genuinely cannot see the defects that matter.
Distribution, 11 kV and LT on poles. Far more line length, far lower value per kilometre, running along roads and through villages and fields with people, trees, buildings and other services all around. Detailed per-structure inspection at 11 kV is rarely economic on the whole network.
Where drone work does pay on distribution is targeted: post-storm patrol of affected feeders, patrol of a specific section after a trip, survey of feeders with a poor reliability record, corridor and encroachment survey on feeders through wooded or built-up areas, and pre-monsoon patrol of the feeders that historically fail. We would rather tell a discom to fly its worst fifty feeders properly than its whole network badly.
How we do it
Step 1: Scope, line data and what you already know
We start with the line: voltage, circuit configuration, structure type, span range, and the section to be covered by structure number or chainage. Every deliverable we produce is referenced to your structure numbering, because your maintenance records, your work orders and your crews use it.
Then what you already know. Previous inspection reports, known defects, trip history for the section, and where the protection system placed the last fault. An inspection that ignores the utility’s own history is an inspection that reports what you already knew and misses what you did not.
Where clearance compliance is in scope, we need conductor type, stringing tension or sag data, design maximum operating temperature and design wind, and your applicable clearance figures. Without those the modelling cannot be done.
Step 2: Airspace, and the BVLOS constraint
We map the line against the Digital Sky airspace map before quoting and return the zone classification by section. Green zone to 120 metres above ground level without prior flight permission, yellow zone with permission and a lower ceiling, red zone not at all. A long line crosses zones, districts and sometimes states.
The constraint that shapes every power line job in India is visual line of sight. Operation beyond visual line of sight requires specific DGCA permission. Without it, a line is inspected in segments with the crew relocating along it, which is entirely workable and is how we plan, and which sets the realistic kilometres per day. Anybody quoting a continuous fifty kilometre inspection mission without a BVLOS approval is quoting something they cannot lawfully fly, and it is worth asking the question of every bidder.
We also arrange the utility’s own permission and, where the line crosses sensitive areas, local intimation. Our aircraft carry active Unique Identification Numbers and our pilots hold DGCA Remote Pilot Certificates.
Step 3: Working near live conductors
Nothing about this service requires an outage, and nothing about it requires anybody to approach a live conductor. The aircraft does, and that governs how we fly.
Electromagnetic interference is real and it is the main technical hazard. The field around an EHV conductor disturbs a drone’s magnetometer, and a compass error near a 400 kV line produces heading drift and erratic flight exactly where you least want it. We calibrate away from the line, approach from a known heading, monitor heading behaviour continuously, and treat unexpected yaw as a reason to withdraw rather than continue. On higher voltages we maintain larger standoff and use zoom optics rather than flying closer.
Minimum approach distance is set by voltage and agreed with the utility before the flight, not judged by the pilot on the day. Detail is obtained optically from a safe distance.
Conductors and earth wires are thin and hard to see against sky, and shield wires above the phase conductors are the ones people forget. Flight paths are planned to approach and depart clear of the conductor plane rather than crossing it.
Where the utility prefers an outage for close work, we plan around their switching. Most of what we do does not need it.
Step 4: Flight planning
Ground sample distance is pixel pitch multiplied by flying height divided by focal length. For corridor and vegetation work at survey resolution, a 20 megapixel one-inch sensor with 5472 pixels across 13.2 millimetres, a pixel pitch of about 2.41 micrometres and an 8.8 millimetre lens at 120 metres gives roughly 3.3 centimetres GSD.
Structure inspection is a different mode entirely. It is close-range work with zoom optics at fixed standoff, capturing each structure from multiple positions and angles, and the deliverable is a consistent image set per structure rather than a photogrammetric block. On a detailed tower inspection the image count per structure runs into the tens, and it is the same set of positions on every structure so that this year’s images compare against last year’s.
Corridor mapping, where a photogrammetric or LiDAR product is required, is planned as a multi-strip corridor with cross-strips, for the same drift reasons set out on our road and railway survey pages.
Step 5: Capture
Light matters for defect detection. Flat overcast light is actually better for inspecting hardware than harsh sun, because deep shadow inside a tower structure hides exactly the bolted connections you want to see. For thermal, the reverse applies in a different way: solar loading on components confuses thermal interpretation, so thermal patrol is often flown early morning or after dark where the utility permits night operation, and always with load recorded.
Wind limits close-range work near structures more than it limits open corridor flying, because standoff tolerance is smaller.
Access along a line corridor is usually the day’s constraint. Take-off points need to be reachable, and on a line crossing fields, canals and scrub that means identifying them in advance rather than finding them.
The crew logs flights, structures covered, times, load conditions where relevant, and anything observed but outside scope.
Step 6: Processing and defect reporting
The deliverable on an inspection job is not imagery, it is a defect list.
Images are organised by structure and by component position, so every image has an address: which tower, which leg, which cross arm, which phase. Defects are recorded against that address, classified by type, and assigned a severity against the utility’s own categorisation where one exists, or a simple three-level scheme where it does not.
We report what we observe and how confident we are in it. An insulator that is clearly shattered is a finding. A discolouration that may be contamination or may be a cracked shed is a flag for closer examination, and we label it as such rather than resolving the ambiguity in whichever direction makes the report look more decisive.
Thermal findings carry the measured temperature rise, the reference point used, the load at the time and the ambient conditions.
Where LiDAR and clearance modelling are in scope, the output is a span-by-span assessment at design condition, with infringing spans listed and located.
Step 7: Delivery
Defect list as a structured table your maintenance system can ingest, with imagery linked. Corridor products as GIS layers. A summary report with prioritisation. Everything referenced to your structure numbering.
On a recurring programme the format stays identical between cycles, so year two compares against year one without anybody reformatting anything.
Accuracy and what it means here
Two different accuracy questions arise on this service and they should not be confused.
Positional accuracy on corridor and clearance work follows the usual photogrammetric relationship: horizontal RMSE typically one to two times GSD, vertical two to three times. At 5 cm GSD that is roughly 5 to 10 centimetres horizontal and 10 to 15 centimetres vertical, measured from withheld checkpoints. LiDAR-derived conductor positions are better than photogrammetric ones and the figures depend on the sensor and the control.
Detection reliability is the accuracy question that actually matters on an inspection, and it is not the same thing. Whether a defect is found depends on resolution at the component, viewing angle, lighting, and whether the defect is externally visible at all. A missing bolt on the far side of a member photographed from one side is not detected, however accurate the geolocation is.
This is why our structure inspection protocol captures from multiple fixed positions rather than a single pass, and why we would rather tell you that a structure was covered from three angles than imply that a single orbit constitutes a complete inspection. No aerial inspection is exhaustive, and a report that implies otherwise is doing you harm.
Deliverables and output file formats
| Deliverable | What it is | Format |
|---|---|---|
| Defect register | Findings by structure and component, classified and prioritised, with imagery linked | XLSX, PDF |
| Structure image set | Consistent multi-angle imagery per structure, indexed by structure number and component | JPEG |
| Thermal findings report | Hotspots with temperature rise, reference, load and ambient conditions recorded | PDF, XLSX, TIFF |
| Corridor orthomosaic | Georeferenced image of the right of way | TIFF (GeoTIFF), JPEG |
| Digital terrain and surface models | Ground and canopy along the corridor | TIFF |
| 3D point cloud | Conductors, structures, ground and vegetation, classified | LAS, LAZ |
| Conductor geometry and sag data | Measured conductor positions by span, with survey temperature recorded | XLSX, DXF, LAS |
| Clearance assessment | Span-by-span clearance at design maximum temperature and design wind, infringements listed | PDF, XLSX, SHP |
| Vegetation encroachment layer | Encroaching growth by span, positioned, with height and clearance | SHP, PDF, XLSX |
| Right of way encroachment layer | Structures, occupation and activity under and beside the line, dated | SHP, DXF, PDF |
| Storm damage assessment | Damaged spans and structures listed, with damage type, imagery and access condition | PDF, XLSX, SHP |
| Tower and route inventory | Structure positions, types, and route alignment | SHP, DXF, KML |
| Corridor survey for new lines | Terrain, land use, crossings and constraints along a proposed route | TIFF, SHP, DWG, PDF |
| Change comparison | This cycle against the previous, defects new, persisting and rectified | XLSX, PDF |
| Flight log and survey record | Flights, heights, times, structures covered, drone UIN, pilot licence number |
Where we work
We are based in Ahmedabad and work nationally.
Gujarat. State transmission and distribution networks across Ahmedabad, Gandhinagar, Mehsana, Banaskantha, Sabarkantha, Vadodara, Bharuch, Surat, Rajkot, Jamnagar, Bhavnagar, Junagadh and Kutch. The renewable evacuation corridors out of the Kutch and Banaskantha wind and solar belt are a particular focus, since those lines run through open, sparsely populated country where foot patrol is slow and where a drone covers ground fast.
Rajasthan and the north-west. Evacuation corridors from the Jaisalmer, Bikaner, Jodhpur and Barmer renewable belt, and distribution networks across the state.
Northern India. Punjab, Haryana, Uttar Pradesh, Uttarakhand and Delhi NCR, via Ludhiana, Chandigarh, Meerut, Lucknow, Kanpur and Dehradun.
Central India. Madhya Pradesh and Chhattisgarh, including the generation-heavy corridors around Singrauli, Korba and Raigarh, via Bhopal, Indore, Jabalpur, Raipur and Bilaspur.
Western India. Maharashtra and Goa, via Mumbai, Pune, Nashik, Nagpur and Chandrapur.
Eastern India. West Bengal, Odisha, Jharkhand, Bihar and Assam, via Kolkata, Bhubaneswar, Ranchi, Patna and Guwahati. Cyclone-exposed coastal networks in Odisha and West Bengal are where storm patrol capability matters most.
Southern India. Telangana, Andhra Pradesh, Karnataka and Tamil Nadu, via Hyderabad, Vijayawada, Visakhapatnam, Bengaluru, Chennai and Coimbatore.
What we do not do
We do not locate electrical faults. That is your protection system and your control room.
We do not carry out any live-line work, do not touch conductors or hardware, and do not string, clear or repair anything.
We do not inspect underground cables, substations internals, switchgear, transformers or any enclosed equipment.
We do not certify compliance. We measure and report; the compliance determination is the utility’s, against its own regulations and design data.
We do not assess internal condition of any component. External visual and thermal evidence only.
We do not provide corona and UV detection as a standard service.
We do not fly in red zones, and we do not fly beyond visual line of sight without specific DGCA permission.
We do not report a clearance compliance assessment without the conductor and design data needed to model it.
Frequently asked questions
Can a drone find the fault on a tripped feeder? It can find visible physical damage very quickly once the section is known: a broken conductor, a fallen pole, a tree across the line, a flashed insulator, a burnt jumper. It cannot locate an electrical fault. Narrowing the fault to a section is your protection system’s job; patrolling that section is where the drone saves the time.
How much faster is it than a foot patrol? Substantially, and the honest answer depends on terrain, access and airspace. The bigger difference is not speed but sightline: a drone sees the top of every structure, which a person on the ground looking up cannot.
Do we need an outage? No. Nothing we do requires the line to be de-energised. We work at a minimum approach distance agreed with you and use zoom optics for detail.
Is it safe near a 400 kV line? With the right procedure. The main technical hazard is electromagnetic interference with the aircraft’s compass, which we manage by calibrating away from the line, approaching on a known heading, maintaining standoff and withdrawing if heading behaviour is abnormal. Standoff distances are agreed with the utility before the flight.
Can you check clearance compliance? Yes, where the scope includes sag and tension modelling and you supply conductor type, stringing data, design maximum temperature and design wind. A measured clearance on the survey day is not compliance, because conductor position changes with load, temperature and wind. If a provider offers compliance assessment without asking for that data, they are not doing the modelling.
Can you do thermal? Thermal hotspot detection finds resistive joints and connections before they fail and is one of the most valuable parts of this service. Our sensor capability is stated on this page, and thermal findings are always reported with load and ambient conditions, because a temperature rise is meaningless without them.
How much line can you cover in a day? It depends on whether the job is corridor survey or detailed per-structure inspection, on terrain and access, and on airspace. Detailed structure inspection is measured in structures per day rather than kilometres. We give a rate for your specific line rather than a general figure.
Can you patrol after a cyclone? Yes, as soon as conditions are flyable. We can hold a standing arrangement with utilities that want response capability lined up before the season rather than organised during it.
Is this worth doing on 11 kV distribution? Selectively. Detailed inspection of a whole LT and 11 kV network is rarely economic. Targeted work is: storm patrol, post-trip patrol, poor-performing feeders, wooded corridors and pre-monsoon patrol of feeders that historically fail. We will tell you where the line is.
What do we get at the end? A defect register referenced to your structure numbering, with classified and prioritised findings and imagery linked to each, plus whichever corridor products are in scope. Not a folder of photographs.
Get a quote
Send us the line details: voltage, section by structure number, structure type, approximate length and terrain. Tell us whether you need storm and outage patrol, routine condition inspection, thermal, clearance and vegetation work, or a corridor survey for a new route.
If clearance compliance is in scope, send the conductor type, stringing data and your applicable clearance figures with the enquiry.
We will return the airspace position by section, the achievable rate, the sensors we will fly and a fixed price.