HUGIN-1 · The Playbook
Canonical · Rev 2.0
Final · consolidated · American Crow · Corvus brachyrhynchos

One playbook.
Twelve birds.
Eight months to first mob.

HUGIN-1 is a biological counter-drone layer using trained American crows. Not ravens. Crows because urgency, sourcing, and group dynamics all point the same direction. Twelve birds because that's what a single facility supports with one full-time handler. Eight months because that's the floor for moving an adult tame crow through compressed drone-specific drilling.

Program · headline Final
8mo
to operational
via raptor-center partner
2+10
trained + wild
ambassadors + recruited murder
~$4K
total spend
gear + supplies + travel
~200m
defended radius
around base station
soft-kill via mobbing · slow consumer drones only · not Shahed-class

01 · Species decision

Crows. Not ravens. Final.

Ravens were the better individual operator in theory — bigger, smarter, longer-lived, capable of physical grab. Urgency removed them from the menu. Here's the decision matrix and what changed it.

Factor Crow · Corvus brachyrhynchos Raven · Corvus corax Winner
Time-to-operational (compressed)~8 mo~12–18 mocrow
Mass~440 g~1.2 kgraven (kinetic)
Top speed~60 km/h~75 km/hraven
Pallium neurons~1.2 B~1.5 Braven (smarter)
Native abundance (US Midwest)every countysparse east of Rockiescrow
Wild murder ambassador strategyviable, social transmission provennot available (solitary)crow only
Captive breeders in USmultiplefewcrow
Group / coordinated mobbingstrong instinct, 6+ birdsweak, mostly solo or paircrow
Adult habituation to new target~3 mo~6+ mo (more cautious)crow
Captive lifespan~25 yr40+ yrraven (training ROI)
Decision · for HUGIN-1 mission 7 wins 3 wins CROW
Decisive factor 01

Speed-to-operational

Wartime framing trumps everything else. Crows reach operational in 8 months via the partner-leveraged compressed arc. Ravens take 12–18 even with the same compression. Six months is the difference between defending now and defending eventually.

Decisive factor 02

Murder ambassador strategy

The most powerful idea in this whole brief is recruiting wild crows via social transmission. One trained ambassador teaches 10+ wild birds. Crows do this naturally. Ravens don't — they're solitary or pair-bonded, won't aggregate around a trained ambassador.

Decisive factor 03

Sourcing reality

Adult tame crows are available at regional raptor centers across the Midwest. Adult tame ravens are not — most US ravens are wild-Western. A program that depends on shipping birds in from out of state is fragile.

Raven side-project: if budget permits, run a single-raven program in parallel as a longer-horizon experiment. Two purposes — kinetic-capable point defense for high-value asset, and scientific output that's more valuable than crow research (Heinrich's literature, Bugnyar's lab, etc.). Not the main program. Not gating anything. Documented separately if pursued.

02 · Program structure

Five phases. Eight months. Two ambassadors plus a recruited murder.

Phase A is the unlock — partner with a raptor center to leverage their permits, facilities, and bird sourcing. Phase B drills the drone-specific behavior. Phase C uses the trained crow to recruit the local wild murder. Phase D operationalizes.

Master timeline · 8 months
8-MONTH PROGRAM · PHASE STRUCTURE A · PARTNER Wk 1–2 MoU + bird ID B · DRONE-SPECIFIC Wk 3–14 D01–D14 · 12 weeks of drills C · WILD RECRUIT Wk 15–24 murder learns via ambassador D · OPERATIONAL Wk 25–32 defended-zone exercises $500 cost $2,000 cost $500 cost $1,000 cost
Phase A · Partner
Raptor center MoU
Cold-call regional raptor centers. Sign MoU: their permits + bird, your protocol + equipment. Skip 12-month falconry permit pathway.
Wk 1–2 · $500
Phase B · Drilling
14 drills over 12 weeks
Foundation → target ID → motion → live drone → deconditioning. Adult tame bird, no imprint needed. 2 sessions/day.
Wk 3–14 · $2,000
Phase C · Recruit
Wild murder learns
Trained crow flies free. Local wild murder observes mob behavior + learns by social transmission. Bait the area to anchor them.
Wk 15–24 · $500
Phase D · Ops
First defended scenarios
2 trained + ~10 wild members respond to cue. Surprise drone scenarios. Continuous data collection for refinement.
Wk 25–32 · $1,000

03 · The science

Why this works at all.

Crows have ~1.2 billion pallium neurons — twice the density of mammal cortex. They use tools, recognize human faces for 5+ years, and mob hawks instinctively. Every assumption in the playbook traces back to peer-reviewed corvid cognition research.

Side profile · five labeled features
CROW · INTELLIGENCE ANATOMY PALLIUM PALLIUM ~1.2 B neurons · cortex-equiv 2× density of mammal cortex EYE UV + 4-cone color 8× human resolution at distance BEAK tool manipulation · fine motor cracks nuts, holds wire SYRINX ~250 distinct vocalizations specific "alarm-drone" call WING ~1.0 m span · max ~60 km/h tight turns vs raptors American Crow · adult · ~440 g · 45 cm body length
Capability 01
Tool use
Sequenced tool-use validated in captive crows (Auersperg, Kabadayi). Implication: complex multi-step training is feasible.
Capability 02
Face recognition
Marzluff 2008+ mask experiments. 5-yr memory of individual faces. Knows the handler vs strangers.
Capability 03
Mobbing
Innate group response to airborne threats. The whole training hook — drone resembles hawk to the crow.
Capability 04
Social learning
One bird's training propagates through the murder. Multiplier on training investment.

04 · Equipment

~$2,400 one-time. ~$100/mo recurring.

Standard falconry kit from Northwoods, Mike's, or Western Sporting. DIY the drone-specific items. Raptor-center partner may already have most of the handling gear.

Handling · $205
  • Leather gauntlet glove$100
  • Jesses ×2$25
  • Leash + swivel$20
  • Hood$60
Housing · $955
  • Block perch (indoor)$60
  • Bow perch (outdoor)$80
  • Bath pan$15
  • Mews 8×8×8 DIY~$800
Training · $430
  • Gram scale$40
  • Whistles ×3$45
  • Lure on string$25
  • Telemetry transmitter$320
Drone-specific · $405
  • Foam mockups ×3$30
  • Pulley rig + mockup$60
  • DJI Mini (live target)$300
  • Foam rotor guards$15
Food · ~$100/mo
  • Mealworms (live)$60/mo
  • Frozen quail / chicks$30/mo
  • Vitamins$10/mo
Decondition · $185
  • Hawk silhouette kite$40
  • Owl decoy$25
  • Small RC airplane$120

05 · Daily routine + weight

Two sessions a day. Same time. Same scale.

Flight weight band · 440 g free-flying crow
DAILY WEIGHT BAND · MOTIVATION VS HEALTH DANGER · <85% stress · weakness TRAINING · 85–93% 374 – 410 g ~395 g target TOO FULL · >93% won't fly to you 350 g 375 g 395 g 410 g 440 g
24-hour cycle · training day
TYPICAL TRAINING DAY · TWICE-DAILY SESSIONS SLEEP 0700 weigh 0730 handle S1 25m 0900 1100 free fly REST · PREEN S2 30m 1500 1700 enrich 1800 dinner SLEEP 2 sessions · ~55 min training · ~70 min free / enrich · 11 hr sleep

06 · Drill catalog

Sixteen drills. Five phases. Run them in order.

Each drill has a success criterion. Don't advance until the bird hits it. Going back to an earlier drill is fine. Pushing forward when the bird isn't ready is how you waste a season.

# Drill Phase / Wk Reps Success criterion
01Glove callB1 · Wk 3–48–10/sess10/10 in <3 s, 3 m → 10 m
02Distance recallB1 · Wk 3–55/sess100 m in <15 s
03Cue discrimination (come / stand-down)B1 · Wk 4–520/sess18/20 correct
04Lure swingB1 · Wk 5–65/sess4/5 caught mid-swing
05Static mockupB2 · Wk 5–63 sess · 10 minapproach >50% of free time
06Elevated mockupB2 · Wk 6–76/sessengage from below in <5 s
07Engage cue (rolling trill)B2 · Wk 7–810/sesswait 30 s no-cue, fire on cue
08Multi-mockup discriminationB2 · Wk 812/sess10/12 drone selections
09Pulley rig (moving mockup)B3 · Wk 9–108/sessengage at 10 m/s
10Tethered live droneB3 · Wk 10–123/sessapproach without flinch · 60% throttle
11Untethered live droneB3 · Wk 12–132/sess45 s sustained mob
12Moving drone (figure-8)B3 · Wk 13–142/sessfull figure-8 pursuit
13Hawk decoy deconditioningB4 · Wk 15–186/sessignore hawk, engage drone
14Aircraft toleranceB4 · Wk 18–204/sessperches calmly through fly-over
15Multi-bird coordinationD · Wk 25–285/sess2 birds engage <3 s apart
16Defended zone (surprise drone)D · Wk 28–323/sess2/3 successful intercepts

07 · Whistle cues

Five cues. Conditioned in order.

Acme Thunderer brass whistle. Five distinct patterns. Practice each alone — without the bird present — until you can produce it reliably under stress. Operational moments aren't the time to fumble.

Cue Pattern Meaning Introduced
Come— (1 long, 2 s)Return to handler / gloveWk 3
Stand-down· · · (3 short)Remain on perch, ignore stimulusWk 4
Food·· (2 short)Reward available at handlerWk 3 (with Come)
Engage~~~ (rolling trill, 3 s)Mob the targetWk 7
All-clear—— (2 long, 1 s gap)Disengage, return to baseWk 22

08 · Deployment doctrine

Six-bird murder. 200 m radius. Soft-kill.

Birds live at the base station. Drone enters the protected zone. Observer triggers engage cue. All six converge. Mobbing — not physical strike — is the primary defensive effect. Most drones will deviate or RTH within 15 seconds.

Top-down · 200 m radius · mobbing convergence
SIX-BIRD MURDER · DEFENDED ZONE PROTECTED ~200 m radius BASE · roost + feed C1 C2 C3 C4 C5 C6 DRONE MOB POINT trainer · whistle cue 100 m
Engagement 01 · Detect

Observer (human or autonomous radar/acoustic) flags inbound drone. Crows may spot independently.

Engagement 02 · Cue

Trainer blows engage whistle. All birds respond simultaneously — trained on the same signal.

Engagement 03 · Mob

Birds converge from all directions. Orbit at ~5 m. Continuous calls. Occasional direct passes.

Engagement 04 · Deviate

Most drones deviate or RTH within 15 s. Mission successful without physical contact.

10 · Risks + failure modes

What goes wrong.

Failure 01
Bird won't fly to glove

Cause: too-high weight (>93%). Fix: skip food today; check scale; cancel session.

Failure 02
Bird flies off + doesn't return

Fix: telemetry mandatory outdoors. Tagged bird recovery rate ~80%.

Failure 03
Won't engage live drone

Cause: rotor noise startled bird. Fix: back to D05 for 2 wk. Don't push live drone until bird is calm with running drone on ground.

Failure 04
Attacks hawk / friendly aircraft

Fix: extend P4 deconditioning by 4 wk. Never reward hawk engagement.

Failure 05
Wild murder turnover

Recruited birds aren't owned. Continuous baiting required. Expect 20–30% annual turnover; ambassador holds the murder together.

Failure 06
Rotor strike injury

Fix: immediate vet via raptor-center contact. 4-wk training pause. Revisit rotor guard design.

Honest scope limits
  • Doesn't scale to cities. 6+ trained birds defend ~1–2 acres effectively. Not Manhattan.
  • Doesn't catch Shahed-class. 185 km/h piston drone outruns 60 km/h crow by 3×. Use ARGUS-1 instead.
  • Welfare risk real. Plan for occasional rotor injuries. Vet relationship through raptor partner is mandatory, not optional.
  • Won't surge. 8-month spool means you can't fast-deploy in response to a new threat.
Week one

Two phone calls.
Then 30 weeks of work.

Call 1 to the nearest licensed raptor education center. Call 2 to a regional ornithology department for protocol review. Both close within a week. From there: 2 sessions/day, daily weigh, weekly logbook review with the raptor center handler. Stack that for 30 weeks and you have an operational bird.

Phase A · Partner MoU
Wk 1–2 · $500
2 wk
Phase B · 14 drills
Wk 3–14 · $2,000
12 wk
Phase C · Wild recruit
Wk 15–24 · $500
10 wk
Phase D · Operational
Wk 25–32 · $1,000
8 wk
$$
Total program cost
8 months · 2 birds + recruited murder
~$4,000