FISCHER 26 UAS
TECHNICAL WHITEPAPER

Operational Concept — Fischer Drone

the platform 26 fills three roles simultaneously: persistent ISR (loitering over the area of interest for ~2 hours, detecting and classifying targets with AI), airborne relay (receiving ELRS signals from low-flying FPV drones and forwarding data to Lisa 26 via Starlink), and autonomous EW lock-on against hostile drones (a directional jammer on a pan/tilt mast tracks radar-detected enemy drones after IFF whitelist verification confirms they are not friendly). Fischer 26 carries no sub-drones, FPV drones, or offensive payloads — its only weapon is the directional jammer. Offensive ground engagement is executed by separate FPV strike drones for which Fischer 26 relays commands, not launches.

the platform 26 carries Starlink Mini directly onboard — the same approach publicly reported as used by Ukrainian Baba Yaga and Vampir platforms. This gives it unlimited range (no ELRS distance limitation to ground station) and direct internet connectivity to Lisa 26 regardless of terrain or distance. FPV strike drones are too small to carry Starlink, so they relay through the ISR drone.

Related Chapters

Dual-Role Architecture — ISR and Relay Simultaneously

Fischer 26 is not just a camera drone that happens to carry a radio. The dual-role architecture is intentional: the platform provides persistent ISR coverage (YOLOv8 thermal/visual detection at 300 m AGL cruise) AND serves as an elevated MANET relay node (Silvus StreamCaster at 300 m AGL extending ground-level communication from 3-8 km to 30-50 km). These two functions complement each other perfectly — the drone must be airborne for ISR anyway, and the additional weight of the MANET radio (450 g) is negligible compared to the 8.5 kg total platform weight. The 300 m AGL cruise altitude places Fischer 26 above the effective envelope of AK-family small arms (400-500 m vertical for area targets) while remaining well inside useful ISR resolution. Close-pass engagements at 100-150 m AGL are reserved for Battle Damage Assessment, where the drone enters and exits the small-arms threat envelope briefly (60 s plan budget) for high-resolution imagery of struck targets. See fischer26e.html for the tier-2 variant operating at 500-700 m AGL above the PKM envelope.

Three-tier altitude architecture — how the variants work together

Fischer 26 operates as the lowest tier of a three-level altitude stack. The tiers are coordinated by role, not redundant alternatives:

The tier-3 airframe (Fischer 26E-LE with the antenna cluster) is the most expensive and the one held farthest from the densest threat concentration. Its stand-off station at 1000 m AGL and 3–5 km behind the forward line places it above the practical ceiling of most current-generation Russian FPVs (~600 m effective) while putting its jamming cluster within 5–15 km of tier-1 and tier-2 aircraft operating forward. The cheap, expendable tier-1 Fischer 26 overflies the FLOT and accepts small-arms attrition; the expensive tier-3 Fischer 26E-LE protects it electronically from behind. This is the drone-scale analogue of fixed-wing stand-off strike doctrine — Rafale with SCALP, F-35 with JASSM-ER — adapted to the persistent ISR/EW problem.

The three-tier altitude architecture is also the operational justification for the procurement-tier structure discussed below. Tier A pricing is accepted for the forward Fischer 26 because its expected lifetime is short regardless of hardware provenance; Tier C pricing is accepted for Fischer 26E-LE because its stand-off role preserves the airframe long enough for MIL-qualified components to amortize their cost.

Commercial alternatives serve one function each. DJI Matrice 350 (€12,000): excellent camera system, no MANET relay, no military-grade encryption, DJI cloud dependency, Chinese manufacturer (ITAR/supply chain risk). Silvus MANET repeater drone (€8,000): dedicated relay, no AI inference, no camera. Buying both: €20,000 and two separate aircraft to maintain, charge, and operate. Fischer 26 does both in one platform at any of three procurement tiers — €1,500 Tier A, €5,000 Tier B, or €15,000 Tier C — reducing logistics by 50 percent and operator workload by eliminating coordination between separate ISR and relay assets.

Open Source — CC BY-SA 4.0

Every component specification, ArduPlane parameter, wiring diagram, and software configuration for Fischer 26 is published under Creative Commons Attribution-ShareAlike 4.0. Swedish Armed Forces can build Fischer 26 from these documents without contacting FSG-A, paying any fee, or requesting permission. FOI can modify the design for Swedish-specific requirements. FMV can issue a production tender based on the published specifications. Allied nations in JEF can adopt the platform under the same license.

Why open source for a military system? Because closed-source defense procurement takes 5-10 years from requirement to fielding. The drone threat exists now. Ukrainian units need this capability now. Swedish defense needs to evaluate it now. Publishing openly means any competent engineering team can build a prototype in 2-3 weeks and begin testing. Feedback from multiple builders improves the design faster than any single organization could achieve alone.

Production Scalability

Fischer 26 is designed for production by any competent drone workshop. Most components are commercially available from European, US, or Chinese distributors depending on the tier chosen. No custom-manufactured parts except the fuselage shell (vacuum-bagged fiberglass, producible with €500 of tooling). A workshop producing 5 units per week requires: one CNC router for carbon fiber wing ribs, one vacuum pump and mold for fuselage, one electronics assembly station with soldering equipment. Total workshop investment: €15,000. At €3,000 per Tier B unit and 5 units per week, the workshop generates €780,000 annual revenue. A Ukrainian-pattern Tier A workshop at €1,500 per unit and 20 units per week (the rate Ukrainian volunteer organisations achieve) generates €1,560,000 annual revenue through volume rather than margin — a viable small business or military depot-level production capability at either tier.

Implementation

# Fischer 26 ArduPlane Configuration — GPS-Denied Fixed Wing
# ArduPlane 4.5+ on Pixhawk 6C

# Navigation: NO GPS
param set GPS_TYPE 0              # Disable GPS entirely
param set EK3_SRC1_POSXY 0       # No horizontal position source
param set EK3_SRC1_VELXY 5       # Optical flow for velocity
param set EK3_SRC1_POSZ 1        # Barometer for altitude
param set EK3_SRC1_YAW 1         # Compass (disable if aurora: COMPASS_USE=0)

# Flight envelope
param set ARSPD_FBW_MIN 15       # Min airspeed 15 m/s (54 km/h)
param set ARSPD_FBW_MAX 30       # Max airspeed 30 m/s (108 km/h)
param set ALT_HOLD_RTL 150       # RTL altitude 150m AGL

# Failsafe
param set FS_LONG_ACTN 1         # Long failsafe: RTL
param set FS_SHORT_ACTN 0        # Short failsafe: continue mission
param set FS_BATT_VOLTAGE 20.0   # Battery failsafe at 3.33V/cell (6S)

Material Cost — Three Procurement Tiers

Fischer 26's material cost depends on three procurement decisions, not one: (1) whether the communications stack uses Starlink Mini (only SpaceX sells it, no alternative exists), (2) whether the MANET radio is US-made Silvus (ITAR-controlled) or a Chinese/generic mesh equivalent, and (3) whether peripheral hardware (servos, ESCs, carbon tube, camera modules) is sourced from Chinese retailers (AliExpress, LCSC, 1688.com) or Western distributors. The three tiers below reflect the realistic decision tree:

The three components that cannot be swapped to Chinese

Unlike the Fischer 26E-LE antenna cluster where every component has a Chinese equivalent, Fischer 26 baseline has three irreplaceable Western-sourced items:

1. Starlink Mini — SpaceX is the only supplier. No Chinese LEO satellite broadband constellation exists at comparable coverage or latency. Removing Starlink reduces Fischer 26 from a global-reach ISR relay to a line-of-sight mesh node with 10–30 km range. This is not necessarily wrong for frontline tactical use (Ukraine operates many drones this way), but it is a different capability.
2. Silvus StreamCaster MANET — US-made, ITAR-controlled for military variants. Chinese equivalents (pMDDL, Foxtech, Skydroid H16) exist at 1/5 the price but lack waveform encryption, NSA-type certification, and interoperability with Silvus-equipped ground stations. For a Försvarsmakten integration this is a hard choice; for a Ukrainian brigade it is not a choice at all — Chinese mesh is what they have.
3. Jetson Orin Nano — NVIDIA design, TSMC manufacture. No Chinese SoC delivers 67 TOPS at this thermal envelope. Fallback is running YOLOv8 on the ground station instead of onboard, accepting the 100 ms+ latency for uplinking imagery rather than detections.

Every other Fischer 26 component — airframe, servos, ESC, motor, propeller, battery, camera, flight controller — has a functioning Chinese equivalent at 30–70 % below Western distributor pricing. Tier A reflects this reality. Ukrainian brigade workshops build Tier A-equivalent airframes for €1,000–2,000 per unit in 2026, which is how they field drones at 100+ per week production rates.

Volume pricing — same curve as antenna cluster

For volume-break structure that applies to the airframe hardware, see the antenna cluster page — the same Q1 2026 Western distributor brackets (17 % reduction at 10 units, 40 % at 50, 60 % at 500) are the reference here. Applied to Fischer 26: 50-unit brigade Tier A procurement lands near €700–1,200 per airframe, 500-unit national programme at Tier B lands near €2,000–3,000 hardware per airframe plus subscription costs. The Starlink and Silvus items behave differently: SpaceX subscriptions are flat at €500/month regardless of fleet size, and Silvus units are priced per contract rather than along a published volume curve.

Ukraine 2026 — the empirical innovation hub

The Tier A figures above are not theoretical. They reflect what Ukrainian brigade workshops and volunteer organisations are actually paying in early 2026 to build drones in the hundreds per week. The low cost is not a cosmetic optimisation — it is the reason asymmetric warfare against a materially stronger opponent is sustainable. When an airframe costs under €2,000 in parts, a commander can task it against a €3 million Russian armoured vehicle without the cost ratio making the engagement questionable. When the same airframe costs €15,000, the calculus flips and the engagement must be justified through a longer approval chain. This is not an argument against Tier C hardening; it is an observation that the procurement tier directly shapes which operational decisions are available to a brigade-level commander.

What Ukraine has built since 2022 is a dense ecosystem of hundreds of small manufacturers iterating on drone hardware faster than any Western defence supplier. Airframes that first appear as prototypes in a Kyiv workshop reach serial production within weeks, absorb operator feedback from the front within days, and are discontinued or redesigned within months when Russian countermeasures adapt. This cadence is not possible in the Western defence acquisition system, where a comparable iteration cycle measures in years. The practical implication for Nordic procurement: Ukraine should be treated as the hardware innovation reference for the 2026 drone war, not merely as a consumer of Western donations. Western procurement that ignores this and insists on starting from Western prime-contractor baselines will pay 5–10× more for capability that is already one iteration behind what Ukrainian workshops produce.

The expendability equation — why CPFH and lifetime matter together

A single cost-per-flight-hour figure is misleading without its expected airframe lifetime. Drone classes in Ukrainian service in 2026 span a wide range:

Fischer 26-class aircraft are expected to occupy the upper range of their bracket (30–40 flights rather than the lower 20) specifically because of the countermeasures architecture: frequency-hopping SDR, EMCON modes, and (in the LE variant) two-gimbal engagement of hostile drones that would otherwise force the airframe into low-altitude battle-damage-assessment envelopes. These countermeasures cost money; the justification is that they move Fischer 26 from the 5–15 expendable bracket into the 30–50 sustainable bracket, reducing CPFH by a factor of 3–10×. A hardened drone that flies 40 missions instead of 10 is cheaper per hour even if it costs twice as much to build.

This is why the Tier A / Tier B / Tier C decision is inseparable from the expected lifetime. A Tier A airframe at 40 flights delivers cheaper CPFH than a Tier B airframe at 20 flights; a Tier C airframe at 50 flights is more expensive per hour than a Tier A airframe at 30 flights. The implementing force must project both numbers together — hardware procurement tier and realistic flight lifetime given the expected threat environment — before the CPFH derivation means anything.

Sourcing pattern for Nordic procurement — Ukrainian OEM with Western specification

A sourcing pattern Nordic implementing forces should evaluate: contract Ukrainian manufacturers to build airframes to a Western-authored specification. The implementing force provides the drawings, materials list, acceptance test, and quality gates. The Ukrainian manufacturer supplies volume production capability at Tier A pricing plus the implementation experience of multiple combat iterations. The result is a drone that is Western-specified and acceptance-tested but Ukrainian-built, at 30–50 % of Western-prime-contractor cost.

This pattern is not hypothetical — it is how multiple Western volunteer organisations already fund Ukrainian production at scale. For state-level procurement the obstacles are administrative, not technical: the Nordic procurement authority must accept that the origin-of-manufacture is Ukraine rather than the traditional Western defence prime, and must work through export-control frameworks that currently favour the other direction.

The export-control direction is the harder constraint. Ukrainian-manufactured drones cannot today be re-exported from Ukraine to a third country (including Nordic buyers) without navigating Ukrainian wartime export controls, insurance for equipment in active conflict zones, and the political signal of a Western state buying from a conflict party. Some of these obstacles are genuine; others are inherited from pre-2022 assumptions about arms export directions and could be addressed by bilateral agreements that do not yet exist. An implementing force pursuing this pattern should expect a 12–18 month lead time to establish the procurement framework before first deliveries — which is comparable to the lead time for new Western prime-contractor contracts but with much higher iteration velocity once the framework exists.

The alternative pattern — Tier B or Tier C hardware sourced from established Western suppliers — remains valid for forces that prioritise supply-chain provenance and existing procurement process over unit cost and iteration speed. There is no single right answer; the tier choice depends on operational doctrine, threat expectations, and political constraints specific to each implementing force. What is not defensible is assuming Tier A hardware is automatically inferior to Tier B or Tier C. It is measurably different, but for many tactical roles it is the tier that has been validated in active combat while Western alternatives remain validated primarily against expected threat models in test ranges.

Cost-Per-Flight-Hour Derivation

CPFH depends on both procurement tier and realistic flight lifetime. The derivation below uses 50 flights as a default for continuity with earlier documents; this figure is defensible for Fischer 26-class ISR/EW drones in the 2026 Ukrainian threat environment when Tier B or Tier C countermeasures are present. For Tier A airframes flown in contested airspace without full countermeasures, 20–30 flights is a more honest expectation, which the calculator below allows the reader to vary.

CPFH = (C_airframe / N_flights) + C_energy + C_maintenance

Where:
    C_airframe      = total acquisition cost (tier-dependent)
    N_flights       = expected flights before retirement (20-50 depending on tier
                      and threat environment; Ukraine 2026 empirical data)
    C_energy        = energy cost per flight (EUR 0.10 for 710 Wh LiPo at Swedish
                      grid rate of EUR 0.14/kWh)
    C_maintenance   = pro-rated servicing (EUR 5 per flight: servo replacement
                      every 20 flights, battery replacement every 200 cycles,
                      airframe repair from minor damage)

Substituting for Fischer 26 Tier B baseline (EUR 3,000 subset, 50 flights):
    CPFH = (3000 / 50) + 0.10 + 5 = EUR 65 per 1.6-hour flight
    CPFH_hour = 65 / 1.6 = EUR 40.6 per flight hour

Tier A equivalent (EUR 1,500 hardware, 30 flights — contested airspace):
    CPFH = (1500 / 30) + 0.10 + 5 = EUR 55 per 1.6-hour flight
    CPFH_hour = 55 / 1.6 = EUR 34.4 per flight hour

Tier A equivalent (EUR 1,500 hardware, 40 flights — with countermeasures):
    CPFH = (1500 / 40) + 0.10 + 5 = EUR 42 per 1.6-hour flight
    CPFH_hour = 42 / 1.6 = EUR 26.4 per flight hour

Tier C equivalent (EUR 15,000 hardware, 50 flights):
    CPFH = (15000 / 50) + 0.10 + 5 = EUR 305 per 1.6-hour flight
    CPFH_hour = 305 / 1.6 = EUR 190.6 per flight hour

Worked example — manned ISR comparison. The Swedish Armed Forces operates AJS 37 Viggen reconnaissance sorties at published cost approximately EUR 12,000 per flight hour. Even at Tier C MIL-qualified pricing (EUR 190/hr), Fischer 26 remains 63× cheaper per flight hour. At Tier A with countermeasures (EUR 26/hr), the ratio grows to 455×. This is why Ukrainian brigades can justify operating drones continuously: the Tier A CPFH is dominated by maintenance (EUR 5) rather than amortization (EUR 37.5 at 40 flights), which means flying more does not significantly increase cost per hour.

Verification Code — Cost Calculator

# fischer26_cpfh.py — Cost per flight hour calculator, tier-aware
# Verifies the numbers in the specbox above

def cost_per_flight_hour(airframe_eur, flights_per_lifetime,
                        energy_eur_per_flight=0.10,
                        maintenance_eur_per_flight=5.0,
                        mission_hours=1.6):
    """Return cost in EUR per flight hour."""
    amortization = airframe_eur / flights_per_lifetime
    per_flight = amortization + energy_eur_per_flight + maintenance_eur_per_flight
    return per_flight / mission_hours

def manned_ratio(drone_cpfh, manned_cpfh=12000):
    return manned_cpfh / drone_cpfh

# Three-tier derivation with realistic lifetime per tier
configs = [
    ("Tier A, 30 flights (contested)",       1500, 30),
    ("Tier A, 40 flights (countermeasures)", 1500, 40),
    ("Tier B, 50 flights (Western baseline)",3000, 50),
    ("Tier C, 50 flights (MIL-qualified)",  15000, 50),
    ("Fischer 26E Tier B",                   3900, 50),
]
for label, airframe_eur, flights in configs:
    cpfh = cost_per_flight_hour(airframe_eur, flights)
    print(f"{label:45s}: EUR {cpfh:6.2f}/hr, 1/{manned_ratio(cpfh):.0f}")

# Output:
# Tier A, 30 flights (contested)               : EUR  34.44/hr, 1/348
# Tier A, 40 flights (countermeasures)         : EUR  26.56/hr, 1/452
# Tier B, 50 flights (Western baseline)        : EUR  40.62/hr, 1/295
# Tier C, 50 flights (MIL-qualified)           : EUR 190.69/hr, 1/63
# Fischer 26E Tier B                           : EUR  50.62/hr, 1/237

Why This Matters Operationally

Cost-per-flight-hour matters because it changes the operational calculus of ISR tasking. A brigade commander planning a 24-hour operation can authorize continuous Fischer 26 coverage (10 sorties × 1.6 hours = 16 hours of effective coverage) at a total cost of €650. The same commander cannot authorize 16 hours of manned AJS 37 coverage at €192,000 — that cost exceeds most brigade discretionary budgets. The quantitative consequence is that Fischer 26 enables a category of ISR tasking (persistent, continuous, brigade-discretion) that is simply not accessible with manned platforms at any level of the chain of command below the Air Force Joint Operations Center.

The open-source publishing model amplifies this advantage. Every workshop that builds Fischer 26 from the published specifications lowers the per-unit cost by distributing R&D amortization across more units. A hundred workshops each producing five airframes per month produces 500 airframes monthly — a production rate comparable to Ukrainian wartime Furia or Leleka production — without requiring any central procurement action, vendor negotiation, or ITAR compliance review. This distributed production model is itself a strategic capability that closed-source procurement cannot match at any budget level.

Swedish Supply Chain

The dual-role capability distinguishes the platform from purpose-built alternatives. Commercial ISR drones like the DJI Matrice series provide excellent camera systems but lack the communication relay function. Military relay platforms like the Silvus MANET repeater provide connectivity but lack onboard intelligence. The combined approach means one platform serves both functions simultaneously — reducing the number of aircraft needed and the logistical burden on the brigade drone team.

Open the interactive Coverage Planner →

Sources

Parameter sources. ArduPlane documentation — ardupilot.org/plane, official ArduPilot manual. Pixhawk 6C datasheet — Holybro/ProfiCNC. Jetson Orin Nano Super specifications (67 TOPS) — NVIDIA datasheet. Silvus SL5200 MANET — public documentation from Silvus Technologies. Arducam IMX477 — Sony IMX477 datasheet + Arducam. Infiray T2S+ datasheet. Starlink Mini specifications (1.1 kg, Ku-band) — starlink.com. ArduPlane configuration parameters — ardupilot.org/plane documentation. Formal verification: cost-per-flight-hour, endurance, and airframe-lifetime claims are verified in provable_claims.py (proof FISCHER26_CPFH).

Mathematically verified estimates. Build cost €2,500–4,000 is the sum of components at 2025 prices (BOM in repository). Cruise speed 85 km/h is a propulsion-system power calculation for T-Motor MN5212 at 500 W cruise power. 2-hour endurance is 888 Wh (2× 6S 16000 mAh) / 355 W (average consumption) = 2.5 h theoretical, with 20% reserve giving 2 h practical.

Operational estimates — not validated by FSG-A in the field. 30–50 km MANET range at 200 m AGL is a Friis-equation calculation (33 dBm transmitter + 6 dBi antenna + −90 dBm receiver sensitivity), not measured by FSG-A. Comparisons with DJI Matrice 350 (€12,000) and Silvus relay (€8,000) are from public catalogues, not contracts. "5 units per week per workshop" is a design goal, not validated by real production. €15,000 workshop investment is an estimate, not an actual quote. FSG-A has not built Fischer 26 to flight prototype.

External standards and references. ArduPlane documentation (ardupilot.org). Starlink Mini specifications (starlink.com). T-Motor MN5212 datasheets. NATO STANAG 4671 (UAV Airworthiness). NATO STANAG 4609 Ed. 4 (motion imagery metadata for ISR video). ICAO Doc 10019 "Manual on Remotely Piloted Aircraft Systems". CC BY-SA 4.0 — creativecommons.org/licenses/by-sa/4.0/legalcode. Fischer 26 design documentation — FSG-A open specification.