Sir Airffic GoFrom here on, interpretation includes Airffic's perspective. Separate fact from interpretation and apply it to your own company.
Choosing where to compete: attractiveness × right to win
Plot options on two axes: market attractiveness (growth, profit, stickiness) and your right to win (strengths you already have).
Start top-right — attractive and winnable. Fighting giants bottom-left becomes a war of attrition however good your technology is.
- ★ Airspace OS / UTM
- ★ Inspection & survey ops data
- DFR ops (police, fire)
- Delivery ops backbone
- C-UAS integration
- Build autonomy software
- Build sensors or chips
- Manufacture aircraft
Why the airspace OS and operations layer (three reasons)
- Demand is created by regulation: airspace management and operational records are conditions for BVLOS. The more rules open, the more this layer is needed.
- Lock-in works: procedures, approvals and records embed deeply in customer workflows and are hard to replace.
- Brain progress becomes a tailwind: as aircraft get smarter and more numerous, sky-wide coordination gains value. You ride chip and foundation-model progress instead of fighting it.
- Each drone senses and perceives
- The airspace OS merges one shared sky map
- Mission Control, Vision and Routing & Stock use it
Three entry scenarios
A: Start from operations (recommended)
Low risk, compounding
Enter through operations management for inspection, DFR or delivery. Build operations data and approval track record, use it to train perception AI, then expand into the airspace OS.
B: Connect to the brain (partner)
Medium risk, speed
Integrate with widely used foundations such as Auterion, PX4 and NVIDIA to become aircraft-agnostic: the same operations on any airframe.
C: Build the brain (in-house)
High risk, high return
Build perception AI for one job, such as power-line inspection. Requires exclusive data and risks being overtaken by general foundation models.
Five lines you must not cross
- Stay aircraft-agnostic: lean too hard on one maker and you sink with it when regulation or decoupling shifts.
- Decide your distance from defense first: what you will and won't do, documented against export controls and ethics before you start.
- Be transparent about data: where it is stored and who can see it. Trust is the biggest product.
- Never inflate numbers: always attach sources and assumptions to market or performance figures. One exaggeration can erase fifteen years of credibility.
- Co-create with regulators: do not wait for rules; join the dialogue with trial data.
Sir Airffic GoThe harder you attack, the earlier you draw the lines you'll defend. The great founders of lasting eras never crossed theirs.
A 90-day plan
| Days | Do | Output |
|---|---|---|
| 1–30 | Narrow to one job (e.g. power-line inspection). Interview 10 customers about operational effort and safety worries | List of customer pain points and a willingness-to-pay hypothesis |
| 31–60 | Build a minimal operations-management tool that connects to existing aircraft and trial it with 2 customers | Operations records and safety evidence usable in approval applications |
| 61–90 | Pick one task from the operations data that perception AI can automate and test it. Share trial results with the regulator | Evidence for pricing decisions and the next funding plan |
Numbers to track (KPIs)
| KPI | Why it matters |
|---|---|
| Aircraft one pilot can operate at once | Most direct measure of autonomy and operations value |
| Human hours per flight | The reason customers pay (labour savings) |
| Number and scope of approved flights | Proof of crossing the regulatory wall; hard to copy |
| Accumulated flight hours and annotated images | Fuel for perception AI; depth of the data moat |
| Customer retention | Confirms lock-in is real |
What you can start with Airffic today
Manage operations with Airffic Operation, register aircraft with Airffic Aircraft, book take-off and landing sites with Airffic Port, and practise with Airffic Simulator.
For the pilot recruitment and training market, see our sister site, the Pilot Market Institute. Learn flight basics at Airffic Institute.
Frequently asked questions
Where should a newcomer start in the drone AI market?
Rather than competing head-on with incumbents in sensors, chips or aircraft, we think a strong path is to enter through operations and airspace management — where regulation is creating demand and no winner is decided — then connect operations data to perception AI (editors' interpretation).
What are the main risks in a drone business?
Dependence on one aircraft maker, unclear lines with defense use, opaque data handling, unsupported marketing claims and being slow to adapt to regulatory change.