DJI FlightHub 2 update: what’s new?
DJI FlightHub 2 is changing quickly.
For a long time, drone conversations focused mainly on aircraft.
Better cameras.
Longer flight times.
Stronger transmission.
Improved thermal sensors.
Larger payloads.
All of that still matters.
But the biggest shift now is happening around the drone.
The real value is increasingly in the workflow.
How missions are planned.
How drones are dispatched.
How live information is shared.
How data is reviewed.
How reports are created.
How drone operations connect into wider security, inspection, emergency response and survey systems.
The latest FlightHub 2 updates show exactly where DJI is heading.
FlightHub 2 is no longer just a place to view drones on a map.
It is becoming an operational command layer.
The biggest headline: Copilot missions
The most interesting update is DJI FlightHub 2 Copilot.
DJI describes Copilot as a native AI agent inside FlightHub 2 that allows operators to issue instructions using plain language through text or voice. Instead of manually navigating every menu, an operator can describe what they want the drone to do, and Copilot can interpret the request, read drone status, use annotated waypoints, break the task into steps and create an executable action plan.
That is a major change.
It means the operator can move from button-by-button control towards intent-based mission control.
For example, instead of manually building every stage of a task, the operator may be able to say something like:
Fly to the south gate and search for a person wearing a red jacket.
Or:
Fly to waypoint three, hover and take three photos.
Or:
Circle the search area at 50 metres and record video.
DJI’s Copilot manual describes natural language commands such as “Fly to the park entrance and search for a person wearing red”. It also explains that Copilot parses the instruction to identify locations, actions and targets, and that high-risk operations such as take-off or flight require a planning confirmation before execution.
That last point is important.
This is not the drone simply doing whatever someone says.
There is still human confirmation and oversight.
Copilot route drawing
The latest public release notes show another important step: Copilot now supports flight route drawing.
DJI says Copilot can generate waypoint routes by selecting annotation points, framing an area, specifying point locations or using natural language descriptions. The generated route can then be saved to the flight route library or dispatched immediately for execution.
This could be extremely useful for enterprise teams.
In practical terms, it may reduce the time needed to build repeatable routes.
That matters for:
Security patrols
Construction site checks
Infrastructure inspection
Solar farm monitoring
Perimeter review
Incident response
Drone as First Responder
Utilities inspection
Dock 3 workflows
The important point is not just that the route can be created faster.
It is that the route can be created from operational language.
That makes FlightHub 2 easier for non-specialist users to understand, while still keeping the drone operation inside a structured platform.
Why voice and text control matter
Voice and text mission control could sound like a gimmick.
It is not.
In real operations, speed and clarity matter.
A control room operator may not want to dig through multiple screens during an incident.
A security team may need the drone to go to a specific gate, yard or boundary point quickly.
A public safety team may want the drone to search an annotated area and return useful results.
An inspection team may want repeatable data capture without manually rebuilding the same mission again and again.
Natural language control can reduce the gap between the operational need and the drone task.
The operator describes the outcome.
The software helps build the mission.
That is the real value.
Current Copilot capability
DJI’s April 2026 FlightHub 2 update lists current Copilot capabilities as:
Launching and controlling drone missions by text or voice
Reading and interpreting annotated waypoint data
Understanding drone status and translating user intent into actions
Breaking complex tasks into executable steps
The manual also shows support for flight movement commands, Return to Home and land, emergency stop, photo capture, timed or continuous video recording, stopping recording, pano capture, lens switching, camera orientation towards a coordinate, location confirmation, plan confirmation and automatic retrieval of device parameters such as altitude and distance limitations, RTH altitude and failsafe settings.
There are also limits.
DJI’s Copilot manual states that continuous flights between multiple PinPoints are unsupported in a single command, and existing flight task files cannot be executed directly through Copilot remote control; only real-time FlyTo is supported in that context.
That matters.
Copilot is powerful, but it should not be described as fully autonomous mission management without limits.
It is an AI-assisted control layer with human confirmation.
Smart Search in Analyzer
Another important update is Smart Search in FlightHub 2 Analyzer.
DJI’s release notes say Analyzer now supports Smart Search, using large language models to recognise target information automatically. Recognised items can be converted into annotations on the model, and a job report can then be generated.
This is a big deal for inspection and security.
The value of drone data is not just capturing images.
The value is finding what matters in those images.
Smart Search could help teams move from:
“We captured a lot of data.”
to:
“Here are the relevant findings, mapped, annotated and ready for review.”
That is useful for:
Asset inspection
Construction progress
Solar farm inspection
Security review
Infrastructure monitoring
Utilities
Environmental monitoring
Incident documentation
Insurance evidence
Post-storm checks
This is where AI becomes commercially useful.
Not because it sounds clever.
Because it reduces manual review.
VLM Smart Search in workflows
The latest release notes also show that periodic workflows now include VLM Smart Search and Change Detection Pro sub-modules. This means AI analysis tasks can be configured directly into periodic workflows.
That is important for repeat operations.
For example, a Dock 3 could fly a scheduled route across a site.
FlightHub 2 could then support analysis of the collected data as part of the workflow.
That is where automation becomes more than just flight.
The workflow can include:
Fly the route
Capture the data
Search for targets or changes
Create annotations
Support reporting
Help operators review findings
For security and inspection, this is the direction the market is moving.
The drone is only one part of the operation.
The workflow around the drone is what creates the value.
Search In-Place
Trigger-based workflows have also been improved.
DJI says trigger-based workflows now support a Search In-Place action, which can be paired with smart recognition to search for specific targets after the aircraft arrives at a designated location.
This is very relevant for security and Drone as First Responder.
A drone may be sent to an alarm location.
When it arrives, it can search the area rather than simply hover and stream video.
That could support use cases such as:
Search for a person near a fence line
Look for a vehicle at a gate
Check for smoke near a building
Search for a casualty in a defined area
Look for unauthorised activity in a yard
Check whether an alarm is genuine or false
The drone does not replace the operator.
But it can help direct the operator’s attention.
Better trigger-based workflows
DJI’s June 2026 release notes added several trigger-based workflow enhancements.
These include Dock 2 support as an operating device, multimodal recognition that can be activated during flight or on arrival, custom dispatch delay including immediate dispatch after an alert, alert point export to project annotations, and Event API support for pushing dispatch workflow information to third-party systems.
This matters because trigger workflows are where drones start to connect with real-world operations.
A trigger could come from:
A CCTV alert
A perimeter alarm
A control room request
A security event
A public safety incident
A sensor input
A third-party system
The drone response can then be linked to that trigger.
That is powerful.
It turns the drone from a manually launched camera into part of an operational response system.
Better logs and workflow visibility
The release notes also say the workflow list now includes Notification and AI tags, and execution logs support viewing the trigger method and notification results.
That may sound less exciting than Copilot, but it is important.
Serious drone operations need auditability.
Teams need to know:
What triggered the workflow
What action was taken
Whether notifications were sent
Whether AI was involved
What the result was
Whether the workflow needs improving
This is especially important for security, public safety, critical infrastructure and regulated environments.
Multimodal AI in flight routes
DJI says FlightHub 2 now supports Vision-Language Models within flight route planning, expanding recognition beyond fixed categories such as people, vehicles and vessels. This allows the platform to recognise a wider range of objects or scenarios without requiring separate algorithm configuration for every use case.
This could be useful in many sectors.
A security team may want to search for a person in a high-vis jacket.
An inspection team may want to identify damage or missing components.
A construction team may want to find stockpile changes or open excavations.
A utilities team may want to detect abnormal site conditions.
A public safety team may want to search for smoke, fire, vehicles or people in a defined area.
The key point is flexibility.
AI is moving from fixed object categories towards more natural search tasks.
500MP ultra-clear panoramas
FlightHub 2 now supports ultra-clear panoramas up to 500 megapixels.
DJI says the feature captures high-precision partial images using the drone’s telephoto lens and stitches them in the cloud, increasing panoramic image resolution from 100MP to 500MP.
This is useful because many enterprise missions need detail across a wide area.
Examples include:
Large construction sites
Infrastructure projects
Industrial facilities
Ports
Solar farms
Quarries
Public safety scenes
Security monitoring
Remote inspection
Incident documentation
A normal wide image may show the site.
A 500MP panorama may show the detail needed to inspect or report on it.
Change Detection Pro
Change Detection Pro is another important update.
DJI says the Analyzer’s change detection capability has been upgraded with a VLM visual model, allowing non-nadir photos taken from the same waypoint to be analysed for changes. Analyzer also gains a split-screen timeline view, and Change Detection Pro adds area selection so users can focus on where change has occurred rather than reviewing whole datasets manually.
This is very useful for repeat inspection.
The use cases are obvious:
Construction progress
Boundary changes
Solar farm condition
Asset deterioration
Storm damage
Security changes
Earthworks
Infrastructure inspection
Roof inspection
Environmental monitoring
The value is not just seeing change.
It is reducing the time needed to find change.
Resource Library
DJI has consolidated the media library, model library and design files into a single Resource Library. DJI says this makes it faster to find, organise and access project data.
That is useful for teams managing multiple sites and repeat missions.
Good data management becomes more important as drone programmes grow.
A single pilot with one drone may cope with manual folders.
A team operating Dock 3, FlightHub 2 and repeat workflows needs better structure.
Surface area measurement
DJI has added surface area measurement directly inside FlightHub 2.
This allows users to calculate surface area on a model without exporting data to external tools. DJI gives examples including tarpaulin coverage on earthworks and landfills, turf or landscaping costs, and curved or non-standard architectural surfaces.
This is useful for survey, construction and inspection workflows because it reduces the number of software steps.
The less data has to move between platforms, the faster teams can get from capture to decision.
Design file elevation workflows
DJI says FlightHub 2 Analyzer now supports loading design files with a configurable elevation coordinate system. Users can select a reference point and assign a map height, allowing the system to calculate and apply the elevation offset.
That matters for construction and engineering.
Drone data becomes more useful when it can be compared against the design.
For example:
Is the site being built as planned?
Do levels match the design?
Are earthworks progressing correctly?
Where are deviations?
What needs checking on the ground?
This is a practical feature, not just a software enhancement.
Sync 2.0
DJI has replaced the former Cloud Interconnect beta with FlightHub Sync 2.0.
DJI says Sync 2.0 supports project-level file sync, custom flight zone sync, telemetry forwarding through MQTT, live stream forwarding through RTSP and OpenAPI-based management of annotations.
This is important for enterprise customers.
Most serious organisations do not want drone data trapped in a separate system.
They want live streams, telemetry, alerts, files and reports to connect with their existing command centres, VMS platforms, data systems and operational workflows.
Sync 2.0 is part of that direction.
Safesky airspace integration
DJI says FlightHub Sync can integrate external airspace management data, working alongside ADS-B and Remote ID to create a more complete airspace picture. For Europe, DJI says Safesky can be integrated with FlightHub Sync via API to bring in real-time data on non-ADS-B drones and light aircraft nearby.
This is relevant to advanced operations because situational awareness is not only about the ground.
It is also about the airspace.
As remote operations, Dock workflows and BVLOS develop, airspace awareness becomes more important.
One-tap service requests
DJI has also added one-tap service requests inside FlightHub 2.
If a device develops an issue, DJI says operators can submit a service request directly from FlightHub 2, with the system pulling in the relevant device information automatically.
This is a smaller feature, but it matters for fleets.
When drones become operational assets, downtime becomes more expensive.
Anything that reduces maintenance admin is useful.
Supported devices
DJI’s FlightHub 2 support page lists supported DJI Docks including Dock 3, Dock 2 and Dock, and drone platforms including Matrice 400, Matrice 350 RTK, Matrice 30 Series, Matrice 4 Series, Matrice 4D Series and Mavic 3 Enterprise Series, excluding Mavic 3 Multispectral.
That makes FlightHub 2 relevant across a wide range of Enterprise UAV customers.
It is not only a Dock 3 platform.
It can support broader enterprise fleet management, depending on the hardware and workflow.
What this means for public safety
For public safety teams, the update is significant.
Copilot, trigger workflows, Smart Search and AI recognition could help reduce the time between an incident and useful information.
Possible use cases include:
Drone as First Responder
Missing person searches
Incident overwatch
Public order support
Search area scanning
Vehicle tracking support
Fire and smoke checks
Flood assessment
Cordon awareness
Command centre live streaming
The key benefit is not removing the operator.
It is giving the operator better tools.
What this means for security
For security teams, the update makes FlightHub 2 more relevant as part of a control room workflow.
Possible use cases include:
Alarm verification
Perimeter patrols
Gate checks
Industrial estate monitoring
Solar farm security
Port security
Construction site response
Critical infrastructure monitoring
Suspicious vehicle checks
Out-of-hours patrols
Trigger workflows and Search In-Place are especially relevant here.
A drone can be sent to the alert location, then search the area for the target that matters.
What this means for inspection
For inspection teams, the most useful features may be Smart Search, Change Detection Pro, 500MP panoramas, surface area measurement and design file elevation workflows.
These help turn drone data into something more usable.
Possible use cases include:
Roof inspection
Solar inspection
Construction progress
Utilities inspection
Telecoms sites
Industrial plants
Quarries
Earthworks
Flood damage
Infrastructure monitoring
The benefit is not just better data.
It is faster interpretation.
What this means for Dock 3
Dock 3 becomes much more useful when it is paired with better workflows.
A drone-in-a-box system is not just about launching a drone remotely.
The real value is what happens before and after the flight.
What triggered the mission?
What route did it fly?
What did it detect?
What did it report?
Who received the information?
What happened next?
The latest FlightHub 2 updates move DJI closer to that complete workflow.
Dock 3 provides the fixed drone presence.
FlightHub 2 provides the operational layer.
The UK reality
For UK operators, this update is exciting but it does not remove the need for proper authorisation and operational control.
Advanced remote operations, automated dispatch and BVLOS-style workflows still need careful regulatory planning.
The CAA says it is working to address the remaining regulatory challenges to enable authorised BVLOS operators to operate in a scaled and sustainable way, and that safety requires other airspace users, people and property on the ground to be protected.
So the message is:
The technology is moving quickly.
The operation still needs to be legal, safe and properly managed.
Why this update matters
The latest FlightHub 2 updates matter because they change the role of software in drone operations.
FlightHub 2 is becoming less like a dashboard and more like a command system.
It can help teams:
Plan missions faster
Use voice or text to control tasks
Build routes from natural language
Search data using AI
Detect change over time
Trigger drone responses from alerts
Connect drone workflows to third-party systems
Improve reporting
Reduce manual review
Support Dock 3 and remote operations
That is a major step.
The aircraft still matters.
But the workflow is becoming just as important.
Final thoughts
DJI FlightHub 2 is developing quickly.
The latest updates show a clear direction.
Drones are becoming easier to task, easier to integrate and easier to use as part of a wider operational workflow.
Copilot missions are the headline feature because they make drone control more natural.
But the deeper value is in the full platform.
Smart Search.
Route drawing.
Change Detection Pro.
Trigger workflows.
Search In-Place.
500MP panoramas.
Sync 2.0.
Airspace integration.
Analyzer reporting.
Dock 3 workflows.
This is where enterprise drone operations are heading.
Not just better drones.
Better drone systems.
Speak to Enterprise UAV
Enterprise UAV can help organisations understand how DJI FlightHub 2 fits into their drone operations.
We can advise on FlightHub 2, DJI Dock 3, Matrice 4TD, Matrice 400, trigger workflows, Copilot missions, Smart Search, CCTV integration, DFR use cases, inspection workflows, training and CAA authorisation considerations.
Contact Enterprise UAV here:
https://enterpriseuav.co.uk/contact-us/
Shop DJI Enterprise drones and accessories:
https://enterpriseuav.co.uk/shop/
CAA drone guidance:
External sources
DJI FlightHub 2 release notes:
https://fh.dji.com/user-manual/en/release-notes/release-notes-public.html
DJI FlightHub 2 Copilot manual:
https://fh.dji.com/user-manual/en/copilot-assistant/copilot-remote-control.html
DJI FlightHub 2 update: AI Copilot, Safesky and enhancements:
https://enterprise-insights.dji.com/blog/dji-flighthub-2-latest-update-ai-copilot-2026
DJI FlightHub 2 support page:
https://www.dji.com/support/product/flighthub-2
CAA BVLOS policy programme:
https://www.caa.co.uk/drones/drone-regulations/policy-programmes/beyond-visual-line-of-sight-bvlos/
FAQs
What is DJI FlightHub 2 Copilot?
DJI FlightHub 2 Copilot is an AI agent built into FlightHub 2 that allows operators to give drone instructions using plain language through text or voice.
Can FlightHub 2 Copilot create flight routes?
Yes. DJI’s latest release notes say Copilot can generate waypoint routes by selecting annotation points, framing an area, specifying points or using natural language descriptions.
Can Copilot fly a drone automatically?
Copilot can help create and execute drone actions, but DJI’s manual says high-risk operations such as take-off or flight require the user to confirm the plan before execution.
What is Smart Search in FlightHub 2?
Smart Search allows FlightHub 2 Analyzer to use large language models to recognise target information automatically, convert recognised items into annotations and generate a job report.
What is Search In-Place?
Search In-Place is a trigger workflow action that can be paired with smart recognition to search for specific targets after the aircraft arrives at a designated location.
What is Change Detection Pro?
Change Detection Pro helps users compare data from repeat missions and identify changes. DJI says the upgraded version uses a VLM visual model and supports non-nadir photos taken from the same waypoint.
What are 500MP ultra-clear panoramas?
DJI says FlightHub 2 can now create ultra-clear panoramas up to 500MP by capturing high-precision partial images with the telephoto lens and stitching them in the cloud.
Does FlightHub 2 support third-party integration?
Yes. DJI says Sync 2.0 supports project-level file sync, custom flight zone sync, MQTT telemetry forwarding, RTSP live stream forwarding and OpenAPI functions.
Is FlightHub 2 useful with DJI Dock 3?
Yes. FlightHub 2 is particularly relevant for Dock 3 workflows because it supports remote operation, trigger workflows, route planning, AI recognition, data management and integrations.
Do UK operators still need CAA authorisation?
Yes. Advanced remote operations, BVLOS-style workflows or automated deployments may need specific authorisation and operational planning. The CAA says BVLOS development must protect other airspace users and people and property on the ground.