Structural inspection
Looking closely at what is high up
The most common case, and the easiest to decide : there is something to see, and nobody can get up there without rigging. A morning of flying replaces a day of scaffolding, and gives angles no position on the ground can give.
Roofs and coverings
Slopes, ridges, hips, valleys, eaves. Slate, tile, zinc, membrane. Overview survey then the detail points in the same visit.
Stacks and flues
Crowns seen from straight above, condition of the joints, flashings and aprons, lined or unlined flues — the part of the building nobody ever looks at.
Heritage and steeples
Spires, belfries, louvres, long naves. Where the cost of access exceeds the cost of the survey, and where the first reading is often enough to decide what comes next.
Masts and technical sites
Antenna planes, fixings, connectors, labelling, earthing braids. No climbing, so as often as monitoring requires.
Façades and cladding
Complete elevations shot square on, with no perspective distortion, joints and fixings legible to the very top.
Civil structures and industrial sites
Bridge soffits, bearings, silos, flare stacks, structures that are hard to approach. To be scoped together beforehand : every site sets its own access constraints.
The matching shots : photovoltaic case study roof covering chimneys heritage masts
Thermography
Measuring, no longer just showing
The 4TD's sensor is radiometric : every pixel carries a temperature, not just a colour. The difference is between “it is warmer here” and “there are eleven degrees between them” — the second can be argued over, compared from one visit to the next, and put in a report.
Building heat loss
Thermal bridges, missing or slumped insulation, air-tightness defects. In the cold season, when the indoor–outdoor difference works in our favour.
Water trapped under membrane
On a flat roof, water trapped in the insulation gives up its heat more slowly than the rest. The outline reads at the end of the day, and is drawn on the image.
Electrical hot spots
Joints, cabinets, transformers, solar arrays. A failing cell or diode shows from a distance, under load, without shutting anything down.
Search and verification
NIR illumination for night flying, a spotlight, and a thermal signature that does not depend on ambient light.
Existing shots : qualitative readings — measured readings will come with the first 4TD missions.
2D and 3D mapping
The same scene, but measurable
A hundred or so shots resolve into a model where you take distances, areas and volumes. It is a computed deliverable, not a photo : its worth depends entirely on what was done on the ground before take-off.
Orthophoto
A rectified aerial image, to scale and georeferenced, that you measure on directly. The base plan nobody has for their own site.
3D model and point cloud
Textured mesh of a building, a quarry or a site, exportable to the design office's tools.
Volumes and quantities
Stockpiles, heaps, cut and fill, with the reference surface stated explicitly — because it, not the model, decides the figure.
Progress tracking
The same flight plan repeated at regular intervals : two comparable models, and the difference between the two dates rather than an impression.
No mapping shot is published here yet : this page states a capability, the portfolio only shows missions actually flown.
Recurring surveillance
Passing the same spot again, without going back
The 4TD is the “dock” version of the Matrice 4 : IP55, built to stay outdoors and set off on its own along a stored flight plan. What that changes is not image quality, it is frequency — and two identical passes compare, where two visits do not.
Scheduled missions
A route recorded once, replayed identically. Same position, same framing, same hour : the comparison is about the structure, not the flying.
Verification flight
A launch on demand from the supervision desk, with no trip beforehand. Useful where the journey costs more than the flight.
Site monitoring
Quarries, sites, depots, storage yards. The series of images is worth more than any one of them alone.
The dock is an installation
Not an accessory : power, network, a clear location, authorisations specific to beyond-visual-line-of-sight flight. To be studied site by site, before committing to a schedule.
A capability of the aircraft, not an existing installation : the dock is sized with the site, it is not ordered from a catalogue.
The equipment
One aircraft, four sensors
The three focal lengths and the thermal sensor look at the same scene at the same instant : the frame of reference is shared, and a detail seen on the tele drops back onto the wide view without re-registration.
- Aircraft
- DJI Matrice 4TD
- Wide
- 24 mm equiv. · 48 MP · 1/1.3″ · f/1.7
- Mid-tele
- 70 mm equiv. · 48 MP · 1/1.3″ · f/2.8
- Telephoto
- 168 mm equiv. · 48 MP · 1/1.5″ · f/2.8
- Hybrid zoom
- up to 112×
- Thermal
- 640 × 512 native · sensitivity ≤ 50 mK
- IR super-resolution
- up to 1280 × 1024
- Laser rangefinder
- up to 1800 m · ± (0.2 + 0.0015 × D) m
- Illumination
- NIR auxiliary light, 100 m range
- Endurance
- 54 min in flight · 47 min hovering
- Protection
- IP55 · −20 °C to 50 °C
- Automation
- DJI Dock 3 compatible
Manufacturer figures, under test conditions. Real endurance on an inspection runs well below the stated value : you fly slowly, stop, re-frame, and keep a reserve for the return. The rangefinder gives a distance, not a survey coordinate : its accuracy is that of a quick reading, not a topographic survey.
Before the flight
The question that decides the date
None of these jobs starts with the drone. It starts with knowing whether take-off is possible in that place, in that slot — and what to apply for if it is not.
The slot, not only the place
A military area closed in the morning opens in the afternoon. A 150 ft ceiling under an approach path cannot be guessed from any map. A nature reserve appears on no drone map. That is precisely what botic cross-checks, and the tool was born of these flights.
And the diagnosis remains the work of a qualified professional. What is delivered here is images, findings and measurements : enough to look in the right place. The conclusion belongs to the roofer, the engineer or the expert.