Medical drone logistics: cutting laboratory transport from hours to minutes

The structural reason why hospital sample transport exists at all is easy to quantify. According to the Destatis hospital core data for 2022, only 331 of 1,893 German hospitals had their own laboratory medicine department, and only 91 had a transfusion medicine department. For every other house, diagnostics are performed somewhere else: at a partner laboratory, a group laboratory, or the central site of a hospital network. The sample has to travel, and that travel leg is scheduled by a courier plan rather than by clinical urgency.

• Fixed loops: pickup windows are set by the courier route, so a sample drawn just after a departure waits for the next one.
• Ground congestion: urban traffic and roadworks make arrival times variable rather than planned.
• Peak-load batching: a large batch arriving at once creates a queue at the analyser instead of a steady flow.

The effect is that turnaround time is consumed before the sample ever reaches the laboratory. For a ward waiting on a coagulation result or a cross-match, the analytical run may take minutes while the transport leg takes hours. That is the gap short-haul aerial transport addresses, and it is why the business case is almost always built on a specific pair of sites rather than on a general promise of speed.

werob is a manufacturer-independent systems integrator for service robotics. Its drone unit, wedrone, does not fly itself and holds no aviation certificates. It works with approved partner operators and supports hospitals in assessing whether a given route carries a case at all: distance, flight frequency, the laboratory at the receiving end, and the ground handling on both sides.

The strongest German reference is not a pilot. Zollernalb Klinikum operates a route between its Albstadt and Balingen sites of roughly 20 km, flown with the Labfly system from DiAvEn, carrying laboratory samples and blood products. The Luftfahrt-Bundesamt issued the operating approval in August 2024. The operation is classified SAIL III and carries dangerous goods under UN 3373. Test operation started in October 2024, and routine operation has been running since 2025 with around seven flights a day, seven days a week. The remote pilot supervises the flights from Berlin. The aircraft in use, Labfly 1000, is specified with a range of 30 km and a payload of 0.5 to 1 kg.

Those figures define the realistic envelope for this application better than any general claim. This is short-haul, small-payload, high-frequency transport between two fixed sites, not an on-demand air taxi for arbitrary items. The economics reported for the route are equally concrete: around 75,000 EUR per year, compared with 180,000 to 220,000 EUR for the ground solution it replaced.

• Fixed site pair: the route is engineered once, then flown repeatedly. Frequency, not distance, carries the case.
• Small payload: a 0.5 to 1 kg envelope covers sample tubes and blood products, not bulk logistics.
• Continuous flow: samples can be dispatched when they are drawn rather than held for the next scheduled loop.

Other European operations follow the same pattern. In Schleswig-Holstein, Asklepios and MEDILYS have flown a line service between Blomenburg near Selent and Bad Oldesloe since March 2025, roughly 60 km, samples up to 4 kg, also SAIL III. In the UK, Apian operates as a logistics layer inside the NHS with routine operation since 2024, and notably does not fly itself either: it works with Wing and Matternet. A more detailed comparison of what is genuinely in service across Europe is collected in our overview of what actually flies in Europe.

For a laboratory director, the transport question is a pre-analytical question. Launch acceleration, motor vibration and turbulence are mechanical stress, and mechanical stress on whole blood means haemolysis risk. Haemolysed samples are not slow, they are unusable, and a route that produces rejected tubes is worse than the courier van it replaced.

There is no general certificate that settles this. Vibration exposure depends on the airframe, the payload mount, the damping, the tube type and the flight profile, and those differ between platforms and routes. The correct handling is therefore procedural rather than declarative: every new route is validated together with the receiving laboratory before it enters routine operation.

• Paired-sample validation: samples from the same draw are split, one set flown and one set transported conventionally, and the results compared for the analytes that matter on that route.
• Analyte scope: haemolysis-sensitive parameters, coagulation and cell counts define the acceptance criteria, agreed with the laboratory rather than assumed.
• Handling rules: the position of the centrifugation step relative to the flight is part of the validated protocol, not a detail left to the ward.
• Re-validation: a change of platform, payload container or flight profile invalidates the earlier result and requires a repeat.

This is also where the integrator role is concrete rather than cosmetic. The operator holds the aviation approval, the laboratory owns the acceptance criteria, and someone has to write the protocol that connects the two and keeps it documented for the accreditation audit. Hospitals that already run automated internal transport will recognise the pattern from ground systems described in our healthcare robotics overview: the vehicle is the easy part, the process interface is the work.

Aerial transport does not create a separate legal category for the cargo. Diagnostic specimens travel as Biological Substance, Category B under UN 3373, with packing instruction P650: a leak-proof primary receptacle, a leak-proof secondary packaging, absorbent material between the two, and a rigid outer packaging carrying the UN 3373 mark. The Zollernalb operation is approved explicitly for dangerous goods under UN 3373, which is what makes it a transport service rather than a demonstration.

Temperature is the second requirement, and it is set by the payload, not by the aircraft. A drone flies in open air with solar load, wind and altitude-dependent ambient temperature, so the thermal envelope has to be produced inside the payload compartment by insulation, phase change material or active cooling, and it has to be evidenced.

The blood side has its own quantitative argument. According to the Paul-Ehrlich-Institut report under section 21 of the German Transfusion Act for 2024, 3,175,723 erythrocyte concentrates were transfused, 3,469,569 were produced, and 4.43 percent were discarded or expired at the user. Daily demand is around 10,000 units. Faster and more reliable redistribution between sites is one of the few levers that acts directly on that discard rate, which is why blood products appear alongside samples in most serious route concepts.

European drone operations run on two instruments: Implementing Regulation (EU) 2019/947 for the operation and Delegated Regulation (EU) 2019/945 for the aircraft and its components. Regulation 2019/947 defines three categories, open, specific and certified. A medical route flown beyond visual line of sight falls into the specific category. In Germany, the competent authority for the operational authorisation is the Luftfahrt-Bundesamt (LBA).

Inside the specific category there are several routes to an authorisation, and they are not equivalent in effort:

• SORA, the Specific Operations Risk Assessment, is the individual risk assessment. Since 29 September 2025, SORA 2.5 is the binding Acceptable Means of Compliance, with a quantitative iGRC model, ten steps, and the former CONOPS replaced by Detailed Operational Information.
• STS-01 and STS-02 are the standard scenarios, usable by declaration but tightly bounded and not applicable to a typical urban clinic route.
• PDRA, a predefined risk assessment, shortens the process where one exists. The existing set is S01, S02, G01, G02 and G03; PDRA-05 to 08 remain in development.
• LUC, the Light UAS Operator Certificate, lets a mature operator self-authorise within a defined scope. It is an operator maturity instrument, not an entry ticket.

The decisive point for hospital planning: there is still no PDRA covering BVLOS flight over populated areas. A clinic route across a built-up area therefore requires the full individual SORA, and such operations typically land at SAIL III to IV, exactly where the Zollernalb and Asklepios routes are classified. Higher SAIL means higher demands on containment, ground risk mitigation and operator organisation. No pauschal processing time should be assumed for the LBA procedure; it depends on the file.

Airspace integration is regulated separately. The U-space framework, Regulation (EU) 2021/664 together with 2021/665 and 2021/666, has applied since 26 January 2023, but implementation is slow. San Salvo in Abruzzo, 307 km², is the first and so far only fully certified operational U-space airspace in the EU, in service since 1 January 2026, and EU-wide there are only three CISP and three USP. EASA has been consulting on a lighter U-space category since 15 July 2026 under NPA 2026-103, which is not law. In practice, current medical routes are authorised without U-space, through the SORA file and airspace coordination. The approval path is set out in more detail in our article on BVLOS approval for medical drones in Germany.

A flight that lands faster but is logged by hand gives back most of what it won. If laboratory staff have to transcribe arrival times or chase a box, the aerial leg becomes an isolated silo bolted onto an otherwise digital process. The transport event has to appear in the systems that already run the workflow, the hospital information system and the laboratory information system, using the interfaces those systems already speak, typically HL7 and increasingly FHIR.

• Dispatch: the specimen IDs are registered when the transport box is sealed, so the flight is tied to samples rather than to a container number.
• In transit: payload temperature and, where the validation requires it, mechanical exposure are recorded continuously and stored with the flight record.
• Arrival: landing triggers a notification in the laboratory system, so the receiving bench knows a priority batch is on the way from the landing site.
• Audit trail: the complete record, sample, flight, temperature trace, handover, is retrievable as one object for accreditation purposes.

Two practical constraints usually decide how much of this is achievable. The first is that the flight data originates in the partner operator's dispatch system, which is a separate organisation with its own software, so the interface is an integration project with a contractual dimension, not a plugin. The second is that the landing site is a physical location in or near the hospital with its own access, weather protection and handover procedure, and the last hundred metres from the landing pad to the analyser are frequently the part nobody planned. Ground transport robots can close that gap, and the relevant equipment classes are compared in our robot catalogue.

Hardware selection is the least difficult part of a medical drone programme, and buying an aircraft is rarely the right move for a hospital in the first place. The operating model matters more than the airframe, because the approval, the remote pilots, the maintenance organisation and the ongoing compliance sit with whoever holds the authorisation. In the German references, the hospital is the customer of a service, not an aviation organisation.

One EASA-context figure from July 2026 sets the expectation correctly: of 496 active BVLOS approvals across 11 member states, only 89 are regarded as economically viable. An approval proves that an operation is safe, not that it pays. What separates the viable cases is a route flown often enough that the fixed cost of approval, staffing and ground infrastructure is spread across enough flights, which is precisely the Zollernalb profile of around seven flights a day, seven days a week.

• Route economics: flights per day, days per week, and the fully loaded cost of the transport it replaces, including the courier contract and the staff time spent handling it.
• Regulatory position: which authorisation route applies, whether an operator already holds a comparable approval, and what a route-specific SORA will demand.
• Laboratory acceptance: validated analytes, rejection criteria and the handling protocol, agreed in writing before the first routine flight.
• Ground reality: landing site, access, weather availability, and the documented fallback when a flight cannot go, because the courier does not disappear on day one.
• Local planning law: a second bottleneck alongside aviation law. In the Manna case in Ireland, the aviation authority had approved the operation while local councils refused the planning permission for the hubs.

The Apian model in the NHS is instructive here because it separates the roles cleanly: Apian organises the healthcare logistics layer and works with Wing and Matternet for the flying. That is also the shape wedrone works in, as the drone unit of werob, without holding aviation certificates of its own. The task on the hospital side is to specify the route, define the acceptance criteria and the interfaces, and select an approved operator who can actually fly it. Whether the case closes is decided by frequency and by the cost of the transport being replaced, and that calculation belongs at the start of the project, not at the end.