Taiwan Already Has 24 Drone Logistics Corridors — It Didn't Take the Wait-for-Regulation Route

🌏 中文版

At the end of the inspection piece I offered a guess: logistics was probably "no real commercial case in Taiwan, and the reason is that BVLOS door."

Wrong again. And this time more interestingly — because the way logistics routes around that door is the mirror image of inspection.

The numbers first: 10 cases, 24 corridors

Per the CAA in May 2026, drone logistics in Taiwan remains at an early stage, but over recent years the CAA has approved 10 logistics cases using a cumulative 24 drone flight corridors.

The distribution and cargo are specific:

This isn't a concept demonstration. Behind it is a six-year policy path with explicit milestones.

Six years, three phases, and it followed the white paper

The Ministry of Transportation's Institute of Transportation (IOT) has run the "Drone Logistics Service Project for Remote Areas" since 2020. Per the IOT's own account, it advanced in three phases:

PoC  Technical verification  2020–2022  Donggang–Liuqiu (Pingtung), Lalashan (Taoyuan)
                             → verify maritime crossing and mountain flight
PoS  Service verification    2023–2025  Kiwit and Dali indigenous communities, Hualien
                             → extended field operation, operational experience, reliability
PoB  Commercial validation   2025 on    With Chunghwa Post and public-private partners
                             → build a sustainable operating model, normalize the service

And that cadence wasn't back-fitted. The IOT's project record notes that the ministry's drone technology industry group set the milestone in the 2021 Transportation Technology Industry White Paper: complete PoC and PoS for remote and offshore areas before 2024, and PoB in 2025. Actual progress matches that five-year-old plan.

The three application modes the project defined are worth recording, because they explain everything that follows:

• Daily service: two-way movement of daily necessities and agricultural produce for remote residents
• Emergency support: time-sensitive items such as medicines and medical equipment
• Disaster response: when a remote area is cut off by disaster, the daily logistics drone immediately converts to relief use

The third is the design core, which the IOT summarizes in one line: "accumulate capacity through commercial operations during peacetime, and deploy immediately for support during disasters."

It routes around regulation the opposite way from inspection

Now the question.

The inspection piece concluded that inspection routes around BVLOS by cutting work into units completable within visual line of sight — one bridge, one line segment, one 2-to-3 km round trip.

Logistics cannot cut that way. Delivering from A to B while you can't see it is the definition of last-mile.

So it took the other route: one corridor at a time, approved case by case.

Twenty-four corridors across 10 approvals — which is precisely the "one waiver at a time" model the industry map criticized when discussing inspection, saying it "prevents repeated fixed-route work from scaling."

That criticism holds for inspection and fails for remote-area logistics, and the difference is the shape of the demand:

The same regime is an obstacle to coverage-shaped demand and a viable path for route-shaped demand. Remote logistics demand was always concentrated on a handful of lines that break when the boat doesn't sail, so 24 corridors isn't a compromise — it's close to the whole thing.

That also explains the corridor locations: mountain communities, offshore islands, remote areas — all places where one route failing means there is no alternative.

Liuqiu: the value isn't being cheap, it's flying when boats don't

July 2026 at Liuqiu demonstrated this more clearly than any planning document.

The situation: Liuqiu shipping was disrupted by Typhoon Bavi, suspended from the afternoon of 9 July through the 12th and resuming only on the 13th; that same afternoon a high wind warning suspended sailings again for the next day. Supplies on the island were running short, and commuters were caught out.

And the commercial validation was already running there. The Pingtung county government's account: as long as wind isn't excessive, a drone reaches the Liuqiu helipad from the Xinyuan shipping centre in a bit over ten minutes.

But the line worth quoting is the honest description from Chiang Chieh-lun, director of NPUST's uncrewed vehicle applications R&D centre:

With current aircraft, above 10 metres per second of wind the flight risk increases, and during a typhoon it may not fly either. But at night, when there are no sailings, if you need to move emergency supplies or medical items, a drone is genuinely useful.

That places the value of logistics drones very precisely: it isn't replacing the boat, it's covering the boat's gaps.

The aircraft NPUST's team is testing carries about 4 to 5 kg and crosses from Yanpu to Liuqiu in about 20 minutes. Four or five kilograms will never beat a vessel's capacity — but during the hours or days with no vessel, it is the only thing still moving.

The county government's fuller framing: when swells prevent a rescue boat from sailing, in the stage before "emergency helicopter evacuation or naval transfer" becomes necessary, a drone balances the logistics team's risk, cost, and timeliness at once.

This is a value logic different from all three previous pieces. Lined up:

Logistics sells marginal availability, not cost advantage. And that value exists only where the alternative is interrupted — offshore islands, mountain communities, typhoon season. Where the alternative is reliable (urban delivery), the same aircraft has almost no footing.

One detail with an image to it: Liuqiu briefly had a private airport in 1975, which closed quickly for poor economics. Fifty years later, that apron has been reactivated as a drone logistics landing site.

The CAA has quantified the threshold for routine operation

From "24 approved corridors" to "routine commercial operation" is still a distance, and the CAA has put numbers on it. Per the same briefing:

On the aircraft — regardless of weight, type inspection of the airframe must be completed (see the inspection tiers in the spec sheet piece). For drones over 25 kg, the aircraft must additionally pass seven system reviews covering propulsion, communications, flight control, and performance, and complete 50 hours of test flight before entering logistics service.

On the airspace — the CAA has defined a standard "cargo corridor" width of 100 metres, and advises planners to avoid airport zones and local government red zones where possible, which substantially reduces air traffic coordination costs.

Several things there are worth pausing on:

One: 50 hours of test flight is a real cost barrier. Against the cash structure computed in the financials piece — assets paid first, revenue collected later — 50 test hours are pure outlay that must complete before any revenue exists. For a small vendor that is a substantive capital barrier.

Two: "avoiding red zones substantially reduces coordination costs" is a genuinely useful sentence. It tells operators that the first variable in corridor planning is airspace class, not shortest path — directly connected to the red/yellow/green zones in the plain-language regulation piece.

Three: the 25 kg line appears again. The crash anatomy piece established that the TTSB's major occurrence threshold is "over 25 kg substantially damaged"; the spec sheet piece established that over 25 kg requires additional physical inspection; and now logistics hangs its seven system reviews and 50 test hours on the same line. 25 kg is the single most consequential boundary in Taiwanese drone regulation, and it has now shown up once in each of four separate regimes.

Three judgments

1. Rather than "is there regulation," ask whether the regime's granularity fits the demand's shape. Taiwan has no general BVLOS rule, but 24 case-approved corridors are nearly sufficient for remote logistics, because the demand was always concentrated on a few fixed routes. The same regime obstructs inspection — the difference isn't regulatory permissiveness, it's whether demand is dispersed or concentrated.
2. Logistics drones sell marginal availability, not cost advantage. Four or five kilograms will never beat a vessel, but when the vessel doesn't sail it is the only option. That means the market exists natively where the alternative gets interrupted, not where the alternative is expensive — two things routinely conflated.
3. From approval to routine, the threshold is computable: type inspection + seven system reviews + 50 test hours + a 100-metre corridor. That isn't a vague "waiting on regulation," it's a list you can cost and schedule. Anyone entering this business should treat it as a project plan, not a grievance.

References

Policy and overall progress

• CNA — CAA: Taiwan has accumulated 24 drone logistics corridors (in Chinese; 10 cases across 24 corridors, locations and cargo, type inspection plus seven system reviews plus 50 test hours for routine operation, the 100-metre cargo corridor standard, expansion of remote medical delivery)
• CNA — IOT's remote-area drone logistics enters commercial operation validation (in Chinese; PoC/PoS/PoB phases and years, the three application modes, the Chunghwa Post partnership)
• Institute of Transportation, MOTC — Drone logistics service verification project (in Chinese; the 2021 white paper milestones, site selection and inspection procedures)
• Institute of Transportation, MOTC — Drone Logistics Service Project for Remote Areas (English account of the same program with phase-by-phase sites)

Pingtung commercial validation (PoB)

• CNA — Pingtung drone logistics commercial validation: cross-strait and campus delivery (in Chinese; the two flagship corridors, participating institutions, May 2026 approval of the CAA drone logistics operations manual)
• Liberty Times — Commercial drone logistics validation begins at Liuqiu and NPUST (in Chinese; the NPUST–7-Eleven corridor and the Yanpu–Liuqiu corridor with Chunghwa Post)
• Liberty Times — Drones resupply an isolated Liuqiu: medicine and supplies tested in poor weather (in Chinese; the July 2026 typhoon suspension timeline, island supply conditions, the ten-minute crossing, positioning when rescue boats cannot sail)
• PTS News — Remote drone logistics begins commercial testing; Liuqiu apron reactivated as a landing site (in Chinese; the 1975 airport closure and apron reactivation, 4–5 kg payload and 20-minute crossing, the 10 m/s wind risk explanation, two further Hualien corridors)

On this site

• The Drone Industry Map: Components, Regulatory Ceilings, and the Non-Chinese Supply Chain Rebuild
• BVLOS in Three Jurisdictions: Taiwan Has No Framework At All
• Inspection Is Taiwan's Furthest-Along Drone Application — Because It Routed Around BVLOS
• After "NT$2M a Year Flying Drones": Agricultural Spraying Has Run a Full Cycle
• Search and Rescue Drones: The One Application Whose ROI Isn't Money — and the Easiest Budget to Cut
• How to Read a Drone Spec Sheet: Which Lines Regulation Turned Into Boundaries
• The Filings Answered What I Assumed Needed an Interview