AgentScout Logo Agent Scout

Shield AI Business Model Deep Dive: How the $12.7B Defense AI Leader Built Autonomous Drone Empire

Shield AI achieved 140% valuation growth to $12.7B by monetizing AI pilot software via B2G contracts. Analysis of Hivemind platform, CCA/LUCAS wins, and defense market positioning.

AgentScout Β· Β· Β· 12 min read
#shield-ai #defense-ai #autonomous-systems #hivemind #drones
Analyzing Data Nodes...
SIG_CONF:CALCULATING
Verified Sources

TL;DR

Shield AI achieved a 140% valuation increase to $12.7 billion by commercializing AI pilot software for autonomous drones. Unlike traditional defense contractors selling hardware, Shield AI monetizes Hivemind software through government contracts with per-unit licensing and integration fees β€” a rare SaaS-like model in defense. The company secured major wins with the U.S. Air Force CCA program and Army LUCAS project, positioning it as the leading Defense AI agent platform. Score: 8.5/10 β€” Strong product-market fit, exceptional growth trajectory, but concentrated customer base presents risk.

Overview

  • Company: Shield AI
  • Founded: 2015-2016 (exact year unconfirmed)
  • Headquarters: San Diego, California
  • Latest Funding: $2 billion Series G (March 2026)
  • Valuation: $12.7 billion (post-money)
  • Revenue: > $540 million (2026 projected)
  • Website: shield.ai

Shield AI develops AI pilot software for autonomous aircraft and drones. The company’s flagship product, Hivemind, enables multi-agent coordination, allowing a single operator to control entire drone swarms. The platform operates without GPS, communication links, or human intervention β€” a technical differentiator in contested environments where adversaries jam signals.

Testing Methodology

This review synthesizes data from 12 sources including official company announcements, Tier A defense media outlets (DefenseScoop, Army Recognition), financial reporting (TechCrunch, Fortune, Bloomberg, Reuters), and public Pentagon budget documents. Revenue figures come from Fortune’s March 2026 report. Contract details were verified across multiple defense industry sources. Competitive analysis draws from parallel funding announcements and publicly disclosed government contract awards.

Product Architecture

Score: 9/10

Shield AI’s product portfolio centers on Hivemind, an AI pilot software platform with three core components:

Hivemind EdgeOS

The foundation layer provides autonomous navigation for air and ground robots. EdgeOS handles:

  • Precise flight control without GPS
  • Real-time sensor fusion
  • Obstacle avoidance in contested environments
  • Mission continuation during communication blackouts

β€œHivemind is the only AI pilot that has flown full-size aircraft without GPS, without communications, and without a human operator on the controls.” β€” Shield AI Official

Networked Collaborative Autonomy

The coordination layer enables multi-agent operations:

  • Single operator controls entire drone swarms
  • Dynamic task reassignment based on battlefield conditions
  • Coordinated strike packages against layered defenses

Platform Integration

Hivemind operates across multiple platforms:

  • Kratos MQM-178 Firejets: Jet-powered aerial targets converted to autonomous strike aircraft
  • VBAT / X-BAT: Shield AI’s proprietary vertical takeoff drones
  • LUCAS: Low-cost one-way attack drones (U.S. Army contract, 2026)
  • Anduril Fury: Autonomous fighter aircraft (U.S. Air Force CCA program)

The platform-agnostic design allows Shield AI to sell software licenses across multiple hardware ecosystems, creating recurring revenue streams from existing government airframe investments.

Business Model

Score: 8/10

Shield AI operates a hybrid B2G model with three revenue streams:

Revenue StreamDescriptionRevenue Quality
Government ContractsDirect Pentagon contracts for AI pilot integrationHigh margin, long-term
Software LicensingPer-unit licensing fees for Hivemind on platformsRecurring, scalable
Integration ServicesCustom integration work for new platformsProject-based

Comparison to Traditional Defense Contractors

DimensionShield AILockheed MartinAnduril
Primary RevenueSoftware licensingHardware salesHardware + software
Asset ModelLight (software-first)Heavy (manufacturing)Mixed
Recurring RevenueHigh (per-unit licenses)Low (one-time sales)Medium
Platform Lock-inLow (multi-platform)High (proprietary)Medium
Development SpeedFast (software cycles)Slow (hardware cycles)Mixed

The software-first approach creates higher gross margins than traditional defense hardware manufacturers. Per-unit licensing generates recurring revenue as the government procures more platforms equipped with Hivemind β€” a model closer to enterprise SaaS than defense contracting.

Key Contracts

ContractCustomerValueStatus
CCA ProgramU.S. Air ForceUndisclosed (major)Won February 2026
LUCAS IntegrationU.S. ArmyUndisclosedWon May 2026
Operation Epic FuryU.S. Central CommandCombat deploymentDecember 2025
Ukraine SupportUkraine MilitaryUndisclosedActive

Market Position

Score: 8/10

Competitive Landscape

CompanyValuationLatest FundingCore ProductRevenue Model
Shield AI$12.7B$2B (Series G)Hivemind AI pilotSoftware licensing
Anduril$30.5B β†’ $60B (rumored)$2.5B + $8B (rumored)Lattice OS + hardwarePlatform ecosystem
Helsing€4.5B (2023)UndisclosedAI for fighters/dronesEuropean defense B2G

Shield AI occupies the software-centric niche in Defense AI. Anduril pursues a full-stack strategy (hardware + software + platform), while Helsing focuses on European defense markets. The U.S. Air Force intentionally selected both Shield AI and Anduril for the CCA program to avoid vendor lock-in β€” validating Shield AI’s platform-agnostic approach.

Competitive Dynamics

Coopetition with Anduril: The CCA program requires Shield AI’s Hivemind to integrate with Anduril’s Fury aircraft. This creates a cooperative relationship where both companies benefit from program success while competing for future contracts.

European vs. U.S. Markets: Helsing dominates European defense AI due to geopolitical sensitivities around U.S. technology in NATO systems. Shield AI’s primary market remains the U.S. Department of Defense.

Financial Performance

Score: 8/10

Growth Metrics

MetricValuePeriodSource
Valuation$12.7BMarch 2026TechCrunch
Previous Valuation$5.3BMarch 2025Fortune
Valuation Growth140% YoY2025-2026Calculated
Revenue> $540M2026 (projected)Fortune
Series G Total$2BMarch 2026TechCrunch

Investor Breakdown

InvestorRoleCommitment
Advent InternationalLead$1B defense tech mandate
JPMorganChaseCo-leadInvestment group
BlackstonePreferred equity$500M preferred shares
OthersEquity + debtSnowpoint, InnovationX, Riot Ventures, Disruptive, Apandion
Debt facilityCredit$250M

The $12.7 billion valuation implies approximately 23.5x forward revenue multiple ($540M revenue), reflecting:

  • 140% YoY growth rate
  • Major contract wins (CCA, LUCAS)
  • Defensive moat (only AI pilot with GPS-denied capability)
  • Expansion potential through software licensing

Budget Context

The U.S. Department of Defense requested $53.6 billion for autonomous systems, drone platforms, and contested logistics in fiscal year 2027, plus $21 billion for munitions, counter-UAS, and Collaborative Combat Aircraft. This creates a $74.6 billion addressable market for Shield AI’s core capabilities.

Technology Differentiation

Score: 9/10

Technical Specifications: LUCAS Platform

SpecificationLUCAS with Hivemind
Unit Cost$35,000
Range1,000-2,000 km
Payload40 lb
ControlSingle operator for entire swarm
DemonstrationFall 2026 (scheduled)
Combat DebutOperation Epic Fury (December 2025)

The $35,000 unit cost creates favorable cost-exchange ratios against expensive air defense systems. A swarm of 100 LUCAS drones costs $3.5 million β€” less than a single Patriot missile interceptor. Saturation attacks can exhaust defender munitions while maintaining favorable economics for the attacker.

Hivemind Core Capabilities

  1. GPS-Denied Operations: Navigation without satellite signals
  2. Communication Blackout Resilience: Mission continuation when jammed
  3. Dynamic Replanning: Real-time route adjustment based on threats
  4. Multi-Agent Coordination: Swarm tactics without central control
  5. Single-Operator Control: One human manages entire swarm

The ability to operate in GPS-denied environments represents a technical moat. Most autonomous systems require satellite navigation or constant communication with operators. Hivemind’s onboard AI enables independence from external infrastructure.

Risks and Challenges

Score: 7/10

Customer Concentration

Shield AI’s revenue derives primarily from U.S. government contracts. Loss of a major program (CCA, LUCAS) would materially impact financials. The company has limited commercial diversification compared to Anduril’s border security and allied nation contracts.

Regulatory Sensitivity

Export controls on autonomous weapons systems restrict international expansion. The International Traffic in Arms Regulations (ITAR) limits Shield AI’s ability to sell to non-U.S. allies, constraining total addressable market.

Technical Dependencies

Hivemind requires integration with hardware platforms Shield AI does not manufacture (LUCAS from SpektreWorks, Fury from Anduril). Hardware failures or platform cancellations directly affect software licensing revenue.

Competitive Pressure

Anduril’s rumored $60 billion valuation indicates substantial capital for competing AI pilot development. Anduril’s full-stack model (owning both hardware and software) may eventually marginalize pure software vendors in defense.

πŸ”Ί Scout Intel: What Others Missed

Confidence: high | Novelty Score: 78/100

While media coverage focuses on Shield AI’s valuation growth and military contracts, the deeper story is the commercialization blueprint for Defense AI software. Traditional defense contractors sell hardware at low margins with long development cycles. Shield AI flipped this model: license AI pilot software per-unit, integrate with existing government-owned platforms, and capture recurring revenue as fleets expand. The $540M revenue on a $12.7B valuation (23.5x multiple) reflects investors betting this SaaS-like model can scale across the $74.6B Pentagon autonomous systems budget.

Key Implication: Shield AI proves software-first companies can compete in defense by licensing AI agents to multiple hardware platforms β€” a model 10x more capital-efficient than building aircraft. Anduril’s full-stack approach requires billions in manufacturing infrastructure; Shield AI achieves similar functionality through software integration contracts, generating higher margins per dollar invested.

Who Should Use This

Best For

  • Defense investors seeking exposure to AI agent platforms with recurring revenue models
  • Pentagon acquisition officers evaluating AI pilot vendors for multi-platform programs
  • Autonomous systems engineers studying swarm coordination architectures
  • Competitive intelligence teams at Lockheed, Raytheon, Northrop tracking startup disruption

Not Ideal For

  • Commercial drone operators β€” Hivemind targets military use cases, not enterprise logistics
  • Short-term traders β€” 23.5x revenue multiple prices in future contract wins
  • International defense buyers β€” ITAR restrictions limit non-U.S. sales

Bottom Line

Shield AI built the first scalable AI pilot licensing model in defense. The company monetizes software through per-unit fees across multiple hardware platforms, creating recurring revenue streams from Pentagon airframe procurement. With 140% YoY valuation growth, major contract wins (CCA, LUCAS), and a technical moat in GPS-denied autonomy, Shield AI emerges as the leading pure-play Defense AI agent platform. The risk: customer concentration and Anduril’s capital advantage could constrain long-term market share.

Score: 8.5/10

Sources

Shield AI Business Model Deep Dive: How the $12.7B Defense AI Leader Built Autonomous Drone Empire

Shield AI achieved 140% valuation growth to $12.7B by monetizing AI pilot software via B2G contracts. Analysis of Hivemind platform, CCA/LUCAS wins, and defense market positioning.

AgentScout Β· Β· Β· 12 min read
#shield-ai #defense-ai #autonomous-systems #hivemind #drones
Analyzing Data Nodes...
SIG_CONF:CALCULATING
Verified Sources

TL;DR

Shield AI achieved a 140% valuation increase to $12.7 billion by commercializing AI pilot software for autonomous drones. Unlike traditional defense contractors selling hardware, Shield AI monetizes Hivemind software through government contracts with per-unit licensing and integration fees β€” a rare SaaS-like model in defense. The company secured major wins with the U.S. Air Force CCA program and Army LUCAS project, positioning it as the leading Defense AI agent platform. Score: 8.5/10 β€” Strong product-market fit, exceptional growth trajectory, but concentrated customer base presents risk.

Overview

  • Company: Shield AI
  • Founded: 2015-2016 (exact year unconfirmed)
  • Headquarters: San Diego, California
  • Latest Funding: $2 billion Series G (March 2026)
  • Valuation: $12.7 billion (post-money)
  • Revenue: > $540 million (2026 projected)
  • Website: shield.ai

Shield AI develops AI pilot software for autonomous aircraft and drones. The company’s flagship product, Hivemind, enables multi-agent coordination, allowing a single operator to control entire drone swarms. The platform operates without GPS, communication links, or human intervention β€” a technical differentiator in contested environments where adversaries jam signals.

Testing Methodology

This review synthesizes data from 12 sources including official company announcements, Tier A defense media outlets (DefenseScoop, Army Recognition), financial reporting (TechCrunch, Fortune, Bloomberg, Reuters), and public Pentagon budget documents. Revenue figures come from Fortune’s March 2026 report. Contract details were verified across multiple defense industry sources. Competitive analysis draws from parallel funding announcements and publicly disclosed government contract awards.

Product Architecture

Score: 9/10

Shield AI’s product portfolio centers on Hivemind, an AI pilot software platform with three core components:

Hivemind EdgeOS

The foundation layer provides autonomous navigation for air and ground robots. EdgeOS handles:

  • Precise flight control without GPS
  • Real-time sensor fusion
  • Obstacle avoidance in contested environments
  • Mission continuation during communication blackouts

β€œHivemind is the only AI pilot that has flown full-size aircraft without GPS, without communications, and without a human operator on the controls.” β€” Shield AI Official

Networked Collaborative Autonomy

The coordination layer enables multi-agent operations:

  • Single operator controls entire drone swarms
  • Dynamic task reassignment based on battlefield conditions
  • Coordinated strike packages against layered defenses

Platform Integration

Hivemind operates across multiple platforms:

  • Kratos MQM-178 Firejets: Jet-powered aerial targets converted to autonomous strike aircraft
  • VBAT / X-BAT: Shield AI’s proprietary vertical takeoff drones
  • LUCAS: Low-cost one-way attack drones (U.S. Army contract, 2026)
  • Anduril Fury: Autonomous fighter aircraft (U.S. Air Force CCA program)

The platform-agnostic design allows Shield AI to sell software licenses across multiple hardware ecosystems, creating recurring revenue streams from existing government airframe investments.

Business Model

Score: 8/10

Shield AI operates a hybrid B2G model with three revenue streams:

Revenue StreamDescriptionRevenue Quality
Government ContractsDirect Pentagon contracts for AI pilot integrationHigh margin, long-term
Software LicensingPer-unit licensing fees for Hivemind on platformsRecurring, scalable
Integration ServicesCustom integration work for new platformsProject-based

Comparison to Traditional Defense Contractors

DimensionShield AILockheed MartinAnduril
Primary RevenueSoftware licensingHardware salesHardware + software
Asset ModelLight (software-first)Heavy (manufacturing)Mixed
Recurring RevenueHigh (per-unit licenses)Low (one-time sales)Medium
Platform Lock-inLow (multi-platform)High (proprietary)Medium
Development SpeedFast (software cycles)Slow (hardware cycles)Mixed

The software-first approach creates higher gross margins than traditional defense hardware manufacturers. Per-unit licensing generates recurring revenue as the government procures more platforms equipped with Hivemind β€” a model closer to enterprise SaaS than defense contracting.

Key Contracts

ContractCustomerValueStatus
CCA ProgramU.S. Air ForceUndisclosed (major)Won February 2026
LUCAS IntegrationU.S. ArmyUndisclosedWon May 2026
Operation Epic FuryU.S. Central CommandCombat deploymentDecember 2025
Ukraine SupportUkraine MilitaryUndisclosedActive

Market Position

Score: 8/10

Competitive Landscape

CompanyValuationLatest FundingCore ProductRevenue Model
Shield AI$12.7B$2B (Series G)Hivemind AI pilotSoftware licensing
Anduril$30.5B β†’ $60B (rumored)$2.5B + $8B (rumored)Lattice OS + hardwarePlatform ecosystem
Helsing€4.5B (2023)UndisclosedAI for fighters/dronesEuropean defense B2G

Shield AI occupies the software-centric niche in Defense AI. Anduril pursues a full-stack strategy (hardware + software + platform), while Helsing focuses on European defense markets. The U.S. Air Force intentionally selected both Shield AI and Anduril for the CCA program to avoid vendor lock-in β€” validating Shield AI’s platform-agnostic approach.

Competitive Dynamics

Coopetition with Anduril: The CCA program requires Shield AI’s Hivemind to integrate with Anduril’s Fury aircraft. This creates a cooperative relationship where both companies benefit from program success while competing for future contracts.

European vs. U.S. Markets: Helsing dominates European defense AI due to geopolitical sensitivities around U.S. technology in NATO systems. Shield AI’s primary market remains the U.S. Department of Defense.

Financial Performance

Score: 8/10

Growth Metrics

MetricValuePeriodSource
Valuation$12.7BMarch 2026TechCrunch
Previous Valuation$5.3BMarch 2025Fortune
Valuation Growth140% YoY2025-2026Calculated
Revenue> $540M2026 (projected)Fortune
Series G Total$2BMarch 2026TechCrunch

Investor Breakdown

InvestorRoleCommitment
Advent InternationalLead$1B defense tech mandate
JPMorganChaseCo-leadInvestment group
BlackstonePreferred equity$500M preferred shares
OthersEquity + debtSnowpoint, InnovationX, Riot Ventures, Disruptive, Apandion
Debt facilityCredit$250M

The $12.7 billion valuation implies approximately 23.5x forward revenue multiple ($540M revenue), reflecting:

  • 140% YoY growth rate
  • Major contract wins (CCA, LUCAS)
  • Defensive moat (only AI pilot with GPS-denied capability)
  • Expansion potential through software licensing

Budget Context

The U.S. Department of Defense requested $53.6 billion for autonomous systems, drone platforms, and contested logistics in fiscal year 2027, plus $21 billion for munitions, counter-UAS, and Collaborative Combat Aircraft. This creates a $74.6 billion addressable market for Shield AI’s core capabilities.

Technology Differentiation

Score: 9/10

Technical Specifications: LUCAS Platform

SpecificationLUCAS with Hivemind
Unit Cost$35,000
Range1,000-2,000 km
Payload40 lb
ControlSingle operator for entire swarm
DemonstrationFall 2026 (scheduled)
Combat DebutOperation Epic Fury (December 2025)

The $35,000 unit cost creates favorable cost-exchange ratios against expensive air defense systems. A swarm of 100 LUCAS drones costs $3.5 million β€” less than a single Patriot missile interceptor. Saturation attacks can exhaust defender munitions while maintaining favorable economics for the attacker.

Hivemind Core Capabilities

  1. GPS-Denied Operations: Navigation without satellite signals
  2. Communication Blackout Resilience: Mission continuation when jammed
  3. Dynamic Replanning: Real-time route adjustment based on threats
  4. Multi-Agent Coordination: Swarm tactics without central control
  5. Single-Operator Control: One human manages entire swarm

The ability to operate in GPS-denied environments represents a technical moat. Most autonomous systems require satellite navigation or constant communication with operators. Hivemind’s onboard AI enables independence from external infrastructure.

Risks and Challenges

Score: 7/10

Customer Concentration

Shield AI’s revenue derives primarily from U.S. government contracts. Loss of a major program (CCA, LUCAS) would materially impact financials. The company has limited commercial diversification compared to Anduril’s border security and allied nation contracts.

Regulatory Sensitivity

Export controls on autonomous weapons systems restrict international expansion. The International Traffic in Arms Regulations (ITAR) limits Shield AI’s ability to sell to non-U.S. allies, constraining total addressable market.

Technical Dependencies

Hivemind requires integration with hardware platforms Shield AI does not manufacture (LUCAS from SpektreWorks, Fury from Anduril). Hardware failures or platform cancellations directly affect software licensing revenue.

Competitive Pressure

Anduril’s rumored $60 billion valuation indicates substantial capital for competing AI pilot development. Anduril’s full-stack model (owning both hardware and software) may eventually marginalize pure software vendors in defense.

πŸ”Ί Scout Intel: What Others Missed

Confidence: high | Novelty Score: 78/100

While media coverage focuses on Shield AI’s valuation growth and military contracts, the deeper story is the commercialization blueprint for Defense AI software. Traditional defense contractors sell hardware at low margins with long development cycles. Shield AI flipped this model: license AI pilot software per-unit, integrate with existing government-owned platforms, and capture recurring revenue as fleets expand. The $540M revenue on a $12.7B valuation (23.5x multiple) reflects investors betting this SaaS-like model can scale across the $74.6B Pentagon autonomous systems budget.

Key Implication: Shield AI proves software-first companies can compete in defense by licensing AI agents to multiple hardware platforms β€” a model 10x more capital-efficient than building aircraft. Anduril’s full-stack approach requires billions in manufacturing infrastructure; Shield AI achieves similar functionality through software integration contracts, generating higher margins per dollar invested.

Who Should Use This

Best For

  • Defense investors seeking exposure to AI agent platforms with recurring revenue models
  • Pentagon acquisition officers evaluating AI pilot vendors for multi-platform programs
  • Autonomous systems engineers studying swarm coordination architectures
  • Competitive intelligence teams at Lockheed, Raytheon, Northrop tracking startup disruption

Not Ideal For

  • Commercial drone operators β€” Hivemind targets military use cases, not enterprise logistics
  • Short-term traders β€” 23.5x revenue multiple prices in future contract wins
  • International defense buyers β€” ITAR restrictions limit non-U.S. sales

Bottom Line

Shield AI built the first scalable AI pilot licensing model in defense. The company monetizes software through per-unit fees across multiple hardware platforms, creating recurring revenue streams from Pentagon airframe procurement. With 140% YoY valuation growth, major contract wins (CCA, LUCAS), and a technical moat in GPS-denied autonomy, Shield AI emerges as the leading pure-play Defense AI agent platform. The risk: customer concentration and Anduril’s capital advantage could constrain long-term market share.

Score: 8.5/10

Sources

fxcoj69l8ypxscmdeyjfdbβ–ˆβ–ˆβ–ˆβ–ˆaznn1732ycg9526jtzlpce8sf1jj1cohβ–‘β–‘β–‘auweevc6wrb7bvegjq5ihux6vhbjhsβ–ˆβ–ˆβ–ˆβ–ˆfivs3to8bz7d2sdvby1d69jazft28ylrβ–‘β–‘β–‘e0mj3rmg4trj3akynjq3nffq1f9nebjeβ–‘β–‘β–‘837o2x7r5w8dm3tt1p5z6n5evlraaqqleβ–ˆβ–ˆβ–ˆβ–ˆ6vxld1nn3qndp36m4g125379qbcv7i9β–‘β–‘β–‘shurmzsjq365mh9vkvpzolf7edyhw0c8β–ˆβ–ˆβ–ˆβ–ˆ2pknifv04sjf6zjo151icnyivcc1kpqaβ–‘β–‘β–‘20ebz0kx02gsv6fd9vgmdsdjvzvzj7gnβ–‘β–‘β–‘f098cc1nkhc27g36l0vc13llxl4gud3lβ–ˆβ–ˆβ–ˆβ–ˆrcpfwa1wfotg46kh34nei6bel95809bpβ–‘β–‘β–‘1s957n2wdbz5ye72swpjzx8kvuq6vhutiβ–‘β–‘β–‘jn6yvh1kn1se3dmmnre9p7wnuwzpzlxrβ–ˆβ–ˆβ–ˆβ–ˆ04aurghqtvudro8d2wukqyojymtpebl6lsβ–ˆβ–ˆβ–ˆβ–ˆ78l7ekt5sduk0mt2dz5t694yb0rb888aaβ–ˆβ–ˆβ–ˆβ–ˆ3ubqpnsm6dzgwx6uob79s9ne801yem56β–‘β–‘β–‘axdbr12b3lpwlbldn5v2id1q5dkpmgh8yβ–ˆβ–ˆβ–ˆβ–ˆbgeqkdj81ymv4tnmxnqouc3lbmjv9xocfβ–ˆβ–ˆβ–ˆβ–ˆg5udlitazhjb5kfjt9sfucyqxc22wv5hβ–‘β–‘β–‘qr1wff1z8s6plmr1b9kbsisu6blb4fβ–‘β–‘β–‘kdzkp9g3jsd8kcjpug6hgitp7s03lsuarβ–ˆβ–ˆβ–ˆβ–ˆ8kovgz8xd4avarno2r702j179jfb07upiiβ–‘β–‘β–‘oxt6p7lt4bg5mrlicx3k5mhyg2pjoyvtvβ–‘β–‘β–‘jjwm31vxc1d5yfsk0jt11knoicy11n6zgβ–ˆβ–ˆβ–ˆβ–ˆlx2ixke7jjdolazt27bcjp4me7u48r1β–‘β–‘β–‘7c06o4vvvvsj5lizpznuddxs2g7y29e49β–‘β–‘β–‘3o0uw0vdxih8faldx8hgqwplgwuv1vxpfβ–‘β–‘β–‘mmr5kdo6x99txg3mmqmylkypnejulqruβ–‘β–‘β–‘0vja8ti4sg8m202h75w0bhcs5xunlni81β–ˆβ–ˆβ–ˆβ–ˆzbrphwl3f0say2rg7wily4qzadd745gb8β–ˆβ–ˆβ–ˆβ–ˆuq30blmw4j83x1bk2s9tep6qsfa121khgβ–ˆβ–ˆβ–ˆβ–ˆav9korrjb6b9re4dzubefvh58pwwu2fβ–ˆβ–ˆβ–ˆβ–ˆ5dpjukf0vki6u50i3v3v490kju96b0g5r9β–‘β–‘β–‘gkrpktdh8njk2x02tpuwar3gmhcf6f3naβ–‘β–‘β–‘fl7wd8cvxpmpcoe8dpk7pep9uuugmw61β–ˆβ–ˆβ–ˆβ–ˆ7c6i7w7mmadc8a33teenn5bi9dxmdggnβ–‘β–‘β–‘1fybl0q9isvhnoh9z98kris8p22z7msmoβ–‘β–‘β–‘3qnqckq4zlim8wp2b5ou9esaugr97rikβ–‘β–‘β–‘8obeelhb37rguf68vtszahgo37ukoiβ–‘β–‘β–‘je6zd1kqfv773y9dnd5xi1vhel3cw9qyβ–‘β–‘β–‘72g7cpv1n3w6d96clbfa5iew465cownriβ–ˆβ–ˆβ–ˆβ–ˆxgmyb4lj8t7pno79yzaryi0xefamycby3β–‘β–‘β–‘d1e1jd946nvfprcaawv9btm4xz9nsdheβ–‘β–‘β–‘awxqgv5p3f64mcr96wj2db65hflsgui35β–‘β–‘β–‘16i10pp2l5ei2bw29zo7k9i2rv0dnaih0pβ–ˆβ–ˆβ–ˆβ–ˆvv6qxeq8jps8nczftn7ec28ofx1x8fvβ–ˆβ–ˆβ–ˆβ–ˆ62ncisk8vygq8j85aizw36r3i9o8seu7β–ˆβ–ˆβ–ˆβ–ˆvho8q887vght6ppeb01os3x78559jlaeβ–ˆβ–ˆβ–ˆβ–ˆceryl51gok5mv19enssyvdcoccii9kjβ–‘β–‘β–‘db5tx7w5hj9