map Mission Planner solar_power Solar Inspection

How to Import Missions

Follow this guide to transfer your flight plan from Dronpoint into your DJI drone.

Method A: Enterprise Drones Mavic 3E · Matrice 30/300

Professional controllers with DJI Pilot 2 support direct KML import.

1

Export & Copy

Click Export KML in Dronpoint. Copy the file to your controller's SD card or internal storage.

2

Import in Pilot 2

Open the app, tap Mission FlightImport KML/KMZ. Select your file. The route is ready to fly.

Method B: Consumer Drones Mini 4 · Air 3 · Mavic 3

info DJI Fly App Users:
Consumer drones don't have a direct "Import" button. You must use the Replace Method — it works by substituting the internal mission file with your Dronpoint export.
1

Create a Placeholder Mission

With drone and controller powered on:

  1. Open DJI Fly and go to camera view.
  2. Switch to Waypoint Mode.
  3. Tap once on the map to add one random point.
  4. Save the mission — name it something memorable, e.g. My_Flight_Plan.

This creates a folder on your controller that we'll replace in the next step.

2

Locate the Mission Folder

Connect your DJI RC to your computer via USB (or insert the SD card). Navigate to:

Android / data / dji.go.v5 / files / waypoint

Inside you'll find a folder with a UUID name created just now. Open it — there's a file called template.kmz.

3

Rename & Replace

  1. Download your mission from Dronpoint (Export → KMZ).
  2. Rename your file to exactly template.kmz.
  3. Copy it into the drone folder and confirm Overwrite.
4

Ready to Fly

Disconnect the controller. In DJI Fly, open your saved mission My_Flight_Plan — the single placeholder point will be replaced by your full Dronpoint mission.

lightbulb Pro Tip: Always double-check your Altitude and Return-to-Home settings before pressing Start.
solar_power No Thermal Camera Required

Solar Panel Inspection
with a Standard DJI Drone

80% of solar panel performance issues — micro-cracks, heavy soiling, delamination and bird droppings — are clearly visible in high-resolution RGB photos. This guide shows you exactly how to fly, shoot and analyse the results using Dronpoint AI.

check_circle Why a regular DJI camera works:
A DJI Mini 4 Pro or Air 3 shoots at 48 MP. At 20m altitude that gives you a ground resolution of ~0.4 mm/pixel — enough to see micro-cracks smaller than 0.5mm. Thermal cameras are needed only for detecting internal electrical faults (hot spots). Everything visible on the surface is captured better by RGB.

flight_takeoffPart 1 — Pre-Flight Planning

Optimal shooting conditions

ConditionIdealAvoid
Time of daySolar noon ± 2 hoursEarly morning, late afternoon
SkyOvercast / thin cloudsBright direct sun (glare on glass)
WindUnder 5 m/sGusts over 8 m/s (blur)
PanelsDry, no frostRain / frost on panels
wb_cloudy Overcast is better than sunny.
Direct sunlight creates specular reflection (glare) on the panel glass which obscures surface detail. A thin cloud layer acts as a giant softbox — diffuse, shadow-free lighting is ideal for detecting surface defects.

Flight altitude & coverage

4-8
Panels per photo
20–30%
Front & Side overlap
-70° to -80°
Camera angle (pitch)
Hover
Flight style (no blur)
compress Important Note on Resolution:
To optimize upload speed and API constraints, Dronpoint automatically compresses your 48MP photos to a maximum width of 2000px on your device before analysis. What the human eye can't see on a 2000px image, the AI won't see either. Flying close to the panels is absolutely critical.

cameraPart 2 — Camera Settings

Always shoot in manual mode — auto exposure will change between frames and make panel-to-panel comparison unreliable.

SettingRecommendedWhy
ModeManual (M)Consistent exposure across all frames
ISO100 (always)Minimum noise — critical for crack detection
Shutter speed1/500s or fasterEliminates motion blur at flying speed
Aperturef/2.8 – f/4Sufficient depth of field, maximum light
White balanceFixed (e.g. 5600K Cloudy)Prevents colour shifts between photos
FocusFixed / manual ∞Autofocus hunts at nadir — lock it
FormatJPEG Fine or RAW+JPEGRAW allows exposure correction in post
warning Never use Auto ISO.
If ISO changes between frames (e.g. 100 to 400 when a cloud passes), the AI will flag noise artifacts as defects and produce false positives. Lock ISO to 100 and let shutter speed absorb light changes.

imagePart 3 — Ideal Photos vs. Common Mistakes

Use these visual references to ensure your dataset is ready for AI analysis.

✅ Ideal Capture
[ Add Image: good-shot.jpg ]
  • 4-8 panels clearly visible
  • Gimbal at -70° to -80° (no sky reflection)
  • Perfectly sharp (Hover & Capture)
  • Even, diffuse lighting
❌ Sun Glare
[ Add Image: bad-glare.jpg ]
Shot straight down (-90°) on a sunny day. The white reflection blinds the AI, causing a "Warning" or "Invalid" status.
❌ Altitude Too High
[ Add Image: bad-altitude.jpg ]
Flying too high to capture everything at once. Panels are just tiny pixels. The AI cannot find physical damage here.
❌ Motion Blur
[ Add Image: bad-blur.jpg ]
Flying too fast with a slow shutter speed. Panel edges are smeared. The AI will struggle to separate dirt from damage.

routePart 4 — Flight Pattern

1

Plan a grid in Dronpoint

Switch to Area Mode in Dronpoint Planner. Draw the polygon around the entire solar array with some margin. Set:

  • Altitude: Fly low enough to see 4-8 panels per frame
  • Camera pitch: -70° to -80° (slightly angled)
  • Overlap: 20-30% (high overlap is not needed for AI analysis)
  • Flight style: Hover & Capture (prevents motion blur)
2

Flight direction relative to panel rows

Orient the grid perpendicular to panel rows where possible. The camera then crosses each row from above rather than flying alongside it — giving a full perpendicular view without glare from the panel tilt.

In Dronpoint, use the Angle slider in Area Mode to rotate the grid until it crosses panel rows at ~90°. The auto-optimize button will suggest the best angle automatically.

3

RTH altitude

Set Return to Home altitude at least 10m above the highest obstruction near the array (trees, inverter boxes, fencing). Set this in DJI Fly before starting the mission.

4

Fly the mission

  • Stay nearby and monitor — do not leave the area.
  • Watch the live feed for shadows — refly affected rows if a large cloud passes.
  • After landing, check that all panels were covered before moving on.

auto_awesomePart 5 — What the AI Detects

Dronpoint AI analyses each uploaded photo and classifies it into one of three states:

check_circleOK

Clean surface, no visible cracks, no delamination.

warningWarning

Light soiling, minor shadow, small surface imperfection — monitor.

errorCritical

Visible crack, delamination, heavy droppings, physical damage. Immediate action.

Issue typeSeverityWhat it looks like
Micro-cracksMedium – CriticalFine dark or bright lines, often spider-web pattern
Heavy soilingMedium – HighBrown/grey opaque patches, uneven coverage
Bird droppingsLow – HighWhite/grey spots, often clustered
ShadingLow – MediumDark areas from trees, chimneys or structures
DelaminationHigh – CriticalBubbling, yellowing, or peeling encapsulant layer
Physical damageCriticalShattered glass, dented frame, visible penetration
DiscolorationMediumBrownish/yellowish tint, snail trail pattern
MisalignmentLowPanel visibly rotated or shifted in mounting

cloud_uploadPart 6 — Uploading & Getting the Report

1

Select your photos

Go to Dronpoint Solar and drag your nadir photos into the upload zone. Up to 50 photos per report. Photos are auto-compressed to ~700KB before AI analysis — originals are archived in full resolution.

2

Fill in project details

Enter the client name and address. This appears in the PDF report header. Used only for the executive summary — not for visual analysis.

3

Wait for analysis

Each photo takes ~5–10 seconds. A 10-photo report typically completes in under 2 minutes. Keep the browser tab open — results stream back as each panel is processed.

4

Download the PDF report

Click Download PDF. The report includes: summary statistics, AI-written executive summary, panel-by-panel table with status and confidence score, and recommended next steps. Reports are saved in your account history and can be re-downloaded any time.

infoPart 7 — What RGB Cannot Detect

RGB inspection has real limitations. It is an excellent first-pass screening tool, but does not replace thermal inspection for certified utility-scale maintenance. Be transparent with clients about this distinction.
IssueRGB (Dronpoint)Thermal
Micro-cracks (surface visible)✓ DetectedIndirect (hotspot only)
Soiling & bird droppings✓ Detected✗ Often invisible
Physical / frame damage✓ DetectedPartial
Delamination (bubbling)✓ Detected✓ Detected
Internal diode failure✗ Not visible✓ Hot spot
Bypass diode bypass✗ Not visible✓ Cell pattern
Shading issues✓ Visible in photo✓ Hot spot
solar_power

Ready to run your first inspection?

Upload your nadir drone photos and get a full AI defect report in under 2 minutes.

auto_awesome Go to Solar Report