Industry Research

Appearance

Assaia -- Deep Research Report ​

Last updated: 2026-03-22

1. Company Overview ​

Full name: Assaia International AG HQ: Zurich, Switzerland (with US offices) Founded: 2018 (some sources cite 2015/2017; the Swiss commercial register entity dates to ~2018) Employees: ~70-80 (2025 estimates, 17% YoY growth) Annual revenue: ~$4M (2025 estimate, per Growjo/CBInsights) Website: https://www.assaia.com

Assaia builds AI-powered software that transforms existing airport camera feeds into structured operational data, enabling real-time turnaround monitoring, predictive analytics, safety alerting, and emissions tracking. The company positions itself as the category leader in AI-driven aircraft turnaround management.

2. Founding & Origin Story ​

Founders ​

NameRole (current)Background
Max DiezChairman & Co-FounderUniversity of St. Gallen (MSc Strategy & Intl Management, 2015). Serial entrepreneur -- founded/scaled several B2B software companies in Europe and Silicon Valley. Originally CEO, transitioned to Chairman.
Dmitry ChugreevCTO & Co-FounderSystems architect with 10+ years building complex systems. Previously Team Leader at Veeam Software (2013-2016). Leads Assaia's technical architecture.
Nikolay KobyshevCo-Founder (departed)PhD and MSc in Computer Vision from ETH Zurich. Co-founded Europe's largest hackathon (Hack Zurich) and AI startups Spectando and Assaia. Now co-founder/CPO at Cerrion (Y Combinator, factory production line CV). Provided the core CV/ML expertise at founding.
Christiaan HenCEO (appointed Jan 2024)Former Head of Innovation at Schiphol Group (Amsterdam Airport). Managed terminal operations and airport capacity development at AMS before joining Assaia.

How They Found Aviation ​

The founding team -- with deep computer vision expertise from ETH Zurich -- initially explored CV applications across food processing, document verification, and retail checkout. They ran workshops showcasing the technology to find product-market fit. At one workshop, the head of innovation from Swissport recognized the value: "If this actually worked, and you can produce real-time videos and insight about what is happening in an internal process of an aircraft then it would be super valuable for everyone in the industry." A basic LinkedIn ad for airport turnaround tracking generated a dozen inbound emails from airline VPs within days. Within two months the team pivoted entirely to aviation.

3. Funding History ​

RoundDateAmountLead InvestorNotes
SeedFeb 2022Undisclosed--Early strategic investors
Seed (extension)Jun 2023Undisclosed--Strategic Partner Community members took equity
Series BDec 2025$26.6MArmira Growth (Munich)Oversubscribed. Participation from existing investors

Total raised (pre-Series B): ~$9.4M (per PitchBook) Total raised (inclusive of Series B): ~$36M

Strategic Investors (equity stakes via Strategic Partner Community) ​

  • Greater Toronto Airports Authority (GTAA) -- operator of Toronto Pearson
  • International Airlines Group (IAG) -- parent of British Airways, Iberia, Vueling, Aer Lingus
  • JetBlue Ventures -- CVC arm of JetBlue Airways
  • Aeroporti di Roma (ADR Ventures) -- first continental European SPC member
  • Marubeni Corporation -- Japanese trading conglomerate, co-owner of Swissport Japan
  • Alaska Star Ventures

Lead Investor Profile: Armira Growth ​

Munich-based VC providing EUR 10-100M growth capital to European tech companies. Portfolio includes osapiens (raised $120M from Goldman Sachs, 2024) and Wemolo (visual AI parking). Network of 100+ industry advisors.

4. Leadership Team ​

Executive ​

  • Christiaan Hen -- CEO (Jan 2024 - present)
  • Max Diez -- Chairman & Co-Founder
  • Dmitry Chugreev -- CTO & Co-Founder
  • Jan Willem Kappes -- Chief Commercial Officer (MSc Air Transport Management)
  • Ana Butter -- CIO/CISO (25+ years technology experience)
  • Daria Zakrevskaya -- Chief People Officer
  • Louise Niven -- Marketing Director

Regional/Sales ​

  • Tim Toerber -- President, Americas (20+ years aviation)
  • Victor Vlaming -- Customer Success, EMEA & APAC (former KLM purser)
  • Jason Fogelman -- VP Business Development, Americas
  • Ethan Gura -- Senior Project Manager, Americas (commercial-rated pilot)

5. ApronAI Platform Architecture ​

Core Philosophy ​

Assaia's platform is designed to work with existing airport camera infrastructure -- VDGS (Visual Docking Guidance Systems) cameras and CCTV surveillance cameras already installed at stands. This is a deliberate strategic choice: airports avoid large upfront hardware investments, and Assaia can deploy on top of sunk-cost infrastructure.

Architecture: Three Technology Pillars ​

  VIDEO FEEDS (existing CCTV / VDGS cameras)
           |
           v
  +-------------------+
  | COMPUTER VISION   |  Object detection, event recognition,
  | (Deep Learning)   |  activity tracking, safety detection
  +-------------------+
           |
           v
  +-------------------+
  | MACHINE LEARNING  |  Predictions (POBT, PRDT), anomaly
  | (Prediction Eng.) |  detection, trend identification
  +-------------------+
           |
           v
  +-------------------+
  | AI OPTIMIZATION   |  Operational recommendations, stand
  | (Decision Layer)  |  allocation, resource optimization
  +-------------------+
           |
           v
  STRUCTURED DATA: timestamps, alerts, predictions, analytics
  APIs, dashboards, mobile alerts, integrations

Camera & Video Infrastructure ​

  • Primary sources: Existing VDGS cameras (e.g., ADB SAFEGATE Safedock A-VDGS with built-in cameras) and CCTV/surveillance cameras
  • Typical setup: ~3 cameras per stand (monitoring stand area + jetty/aerobridge)
  • Heathrow example: 52 cameras across 17 stands in Phase 1 (Terminal 5); 540+ cameras across 116 gates in full rollout
  • Calgary example: phased rollout across 67 gates
  • Munich example: 150 stands initially
  • No proprietary cameras required -- integrates with existing VMS (Video Management Systems)

What the CV System Detects ​

The computer vision pipeline transforms raw video into structured event data. Detected turnaround milestones and events include:

Aircraft Movement:

  • Aircraft arrival at gate (in-block)
  • Aircraft departure from gate (off-block / push-back)
  • Actual Off-Block Time (AOBT) -- measured more accurately than ACARS

Ground Service Equipment (GSE) & Processes:

  • Passenger boarding bridge (PBB/jetbridge) connection/disconnection
  • Passenger stairs arrival/departure
  • Catering trucks on/off stand
  • Fuel truck arrival, fueling start/stop
  • Baggage/cargo loading and unloading
  • Pushback tug connection
  • GPU (Ground Power Unit) connection
  • PCA (Pre-Conditioned Air) connection
  • Wing walker presence

Safety Events (SafetyControl):

  • Aerobridge not fully parked
  • SOP (Standard Operating Procedure) deviations
  • Hazardous situations on the ramp
  • FOD (Foreign Object Debris) detection potential

Emissions Events (EmissionsControl):

  • APU (Auxiliary Power Unit) running status
  • APU usage duration vs. GPU/PCA availability

Output Format:

  • Coordinates, status, behavior, and interactions of objects and people on the apron
  • Automated timestamps for every detected event
  • The system can be trained to recognize any event visible on available video streams

Prediction Engine ​

The prediction engine combines:

  • Real-time CV-generated structured data
  • Historical turnaround data (millions of turns)
  • Live flight data (scheduled times, delays, aircraft type)
  • Weather conditions
  • Airport operational data

Key predictions:

  • POBT (Predicted Off-Block Time) -- continuously updated during turnaround
  • PRDT (Predicted Ready Departure Time)
  • TOBT (Target Off-Block Time) refinement for A-CDM

ML/AI Model Details ​

Assaia does not publicly disclose specific model architectures. Based on the technical profile (ETH Zurich CV lab origins, real-time video processing at scale, object detection + activity recognition):

  • Likely architectures: CNN-based object detection (YOLO family or similar single-shot detectors for real-time performance), activity/event recognition networks, temporal sequence models for prediction
  • Training data: Proprietary dataset of 2M+ monitored turnarounds across 21 airports
  • Edge vs. cloud: Assaia processes video streams; the exact compute topology (edge inference vs. cloud) is not publicly documented
  • Self-learning: BER deployment specifically references "self-learning, AI-based software," suggesting continuous model improvement from operational data

Integration & Data Platform ​

  • APIs: RESTful APIs for machine-to-machine integration
  • Flight Data Output Service for real-time data sharing
  • Integrates with: AODB (Airport Operational Database), A-CDM platforms, SWIM (System-Wide Information Management), airline resource management systems, third-party turn management apps
  • Partner integrations: SITA, Amadeus, INFORM, ADB SAFEGATE
  • Data export: CSV export, API access
  • Auth: SSO (Single Sign-On) support

ADB SAFEGATE Partnership (Key Technical Integration) ​

ADB SAFEGATE's Safedock A-VDGS combines built-in cameras with Assaia's ApronAI image recognition. The VDGS cameras -- already positioned with optimal sight lines to the aircraft stand -- serve as the primary video source for Assaia's CV processing. This is a symbiotic partnership: ADB SAFEGATE gets AI-powered analytics on top of its hardware; Assaia gets pre-installed, well-positioned camera infrastructure at hundreds of airports globally.

6. Product Suite ​

ApronAI (Base Platform) ​

The core visibility platform. Ingests video, generates timestamps, predictions, and alerts.

  • Real-time turnaround event detections
  • Live camera feeds with airport map overview
  • Predicted Off-Block Time (POBT) and Predicted Ready Departure Time (PRDT)
  • Mobile alerting
  • Performance metrics against Precision Time Schedule
  • CSV export and API access

TurnaroundControl ​

Specialized operational interface for airlines and ground handlers.

  • Side-by-side video tiles and process widgets per gate
  • Real-time turnaround progress tracking against plan
  • Exception-based management (alerts only when intervention needed)
  • Zone controller workflow optimization
  • Designed for airline "Zonal Managers" managing multiple gates

StandManager (launched March 2026) ​

AI-powered gate and stand allocation planning. Built in partnership with Transformers Group.

  • Predictive buffer calculation per flight (e.g., likely arrival 10 min early/late)
  • Automated stand reallocation within airport rule sets
  • Planning horizon up to 90 days in advance
  • Can reduce buffer times by up to 5 minutes per flight
  • Unlocks ~5% additional stand capacity at peak periods (~4 additional stands at large airports without building)

SafetyControl ​

Computer vision-powered ramp safety monitoring.

  • Detects SOP deviations (e.g., aerobridge not fully parked)
  • Color-coded safety indicators on live dashboard
  • Mobile alerts for unsafe situations
  • Eliminates need for manual monitoring of all gates
  • Result: 50% reduction in unsafe behaviors reported

EmissionsControl ​

AI-powered APU monitoring for sustainability and regulatory compliance.

  • Uses thermal and acoustic cameras/sensors to detect APU activity
  • Weather-contextualized data (accounts for conditions requiring APU)
  • Tracks APU usage vs. GPU/PCA availability
  • Mobile alerts for violations
  • Copenhagen Airport: first airport worldwide to deploy this at scale (Sept 2025)

Analysis as a Service ​

Consulting-style data offering.

  • Customized data asset creation
  • Benchmark metrics for KPIs
  • Industry-wide Turnaround Benchmark Reports (published annually since 2023)

7. Airport Deployments (21 airports confirmed) ​

Tier 1 Deployments (Flagship/Large Scale) ​

AirportCodeGates/StandsProductsKey ResultYear
London HeathrowLHR116 gates (T2, T3, T5), 540+ camerasApronAI, TurnaroundControlPhase 1: 52 cameras / 17 stands at T5 with BA2019+
Dubai InternationalDXBAll aircraft standsApronAI, TurnaroundControlEmirates: only airline with full-scale AI turnaround mgmt at home hub2025
MunichMUC150 stands initiallyApronAISecond-busiest German airport, 42M pax2025
Toronto PearsonYYZ106 gates (full rollout)ApronAI, TurnaroundControl44% reduction in taxi-in time, 3.4 min avg delay reduction2023+
JFK Terminal 4 (JFKIAT)JFKFull terminalApronAI, TurnaroundControl5 min reduction in ground delays per flightEarly
Seattle-TacomaSEA89 gatesApronAI, TurnaroundControl17% OTP increase (Alaska), 1 min taxi-in reduction2020+
Rome FiumicinoFCO57 gates Phase 1ApronAI, A-CDM integration6 min departure delay reduction, PBB connected 1 min earlier2023+
Los AngelesLAX--ApronAI----
Singapore ChangiSIN--ApronAI----
SydneySYD--ApronAI----

Tier 2 Deployments ​

AirportCodeProductsKey Result
Berlin BrandenburgBERApronAI, TurnaroundControlFirst German airport with AI turnaround; 2.4 min delay reduction
Calgary (YYC)YYCApronAI67 gates phased rollout from 10-gate pilot (Jul 2025)
CopenhagenCPHApronAI, EmissionsControlFirst airport with AI APU emissions monitoring (Sep 2025)
London GatwickLGWApronAIFirst airport worldwide to deploy CV on apron (2018-2019)
NewarkEWRApronAI--
NaplesNAPApronAI--
Halifax StanfieldYHZTurnaroundControlDeployed Jan 2023
CincinnatiCVGApronAI--
Nadi (Fiji)NANApronAI--
Ljubljana (Fraport)LJUApronAI4 min delay reduction; in-house ground handling
Fraport (multiple)--ApronAIExamining rollout across Fraport network

AOBT Accuracy Validation (US Airport) ​

  • 87-gate deployment using existing cameras
  • Study period: May-Sep 2021
  • Found average AOBT inaccuracy of 180 seconds in existing ACARS-based data vs. Assaia CV-measured actual
  • Up to 6 minutes difference between best/worst performing airlines
  • Potential saving: $23/flight or $4.5M/year from better timestamp accuracy alone

8. Airline Customers (460+ airlines exposed to platform) ​

Named Airline Customers/Users ​

Alaska Airlines, United Airlines, British Airways, JetBlue, Air Canada, Delta, American Airlines, Southwest, Hawaiian Airlines, LATAM, Emirates, Virgin Atlantic, Spirit, Frontier, KLM, Air France, Qatar Airways, Fly Dubai, WestJet, Volaris

Headline Airline Results ​

AirlineAirportKey MetricDetail
Alaska AirlinesSEA17% OTP increase12% turnaround time reduction, 25% ground delay reduction, 30% increase in profit/flight, 3.9 min avg delay reduction for delayed flights
United AirlinesNominated hub~2 min avg ground delay reduction$277K savings in Feb 2024; $2.7M/yr at one airport; $169M potential across network
British AirwaysLHR T5First airline to use AI video for punctualityIAGi Accelerator Programme origin; expanding to 116 gates
EmiratesDXBFull-scale AI turnaround mgmtOnly airline in the world operating full AI turnaround system across entire home hub

9. Published Results & Validation ​

2025 Turnaround Benchmark Report (April 2024 - March 2025) ​

  • Scope: 450,000+ AI-enabled turnarounds at 15 airports in Europe and North America
  • Median departure delay reduction: 25% (from 4 min to 3 min)
  • Average delays: stabilized at 11 minutes despite record traffic
  • Gate efficiency improvement: 5%
  • Additional capacity: ~1 extra flight per day per 20 stands
  • Narrowbody turnaround time: 78 min average
  • Narrowbody turns per stand: 4.75/day

Regional Performance ​

Europe:

  • Assaia airports: departure delays 6 minutes lower than 18-minute regional average
  • Savings: ~$600/turnaround, >$70M/year at large hubs

North America:

  • Assaia airports: 11-minute avg delay vs. 12-minute industry average
  • Savings: ~$100/turnaround per minute of delay reduction

2024 Turnaround Benchmark Report ​

  • Ground delays dropped 6% year-over-year
  • Turnaround time decreased 4%
  • 25% improvement in turns per gate (median from 4 to 5 turns/day)

Long-Term Projections (by 2035) ​

  • Major airline potential: $900M in annual benefits from consistently achieving "perfect turns"
  • Large international hub: $300-500M annually

Platform Scale Metrics (as of 2025) ​

  • 2,039,612 turnarounds monitored to date
  • 21 airports deployed
  • 460 airlines using the platform

10. Comparison: Assaia vs. Moonware HALO ​

Fundamental Architectural Difference ​

DimensionAssaia (ApronAI)Moonware (HALO)
Primary data sourceExisting CCTV/VDGS cameras (video)GPS/telematics from mobile devices & GSE
Core functionObservation & analytics (what happened, what's happening, what will happen)Orchestration & dispatching (who should do what, when, where)
Infrastructure requiredExisting cameras (CCTV/VDGS); no new hardware in most casesNo cameras; uses mobile app + telematics devices
Deployment timeWeeks to months (camera mapping, model calibration)As fast as 1 week (software-only, per Moonware)
Value proposition"Digital twin of the turnaround" -- visibility, prediction, A-CDM"Airside OS" -- task dispatching, crew/GSE coordination, ground traffic control
CategoryTurnaround monitoring & analyticsGround operations orchestration & automation

Assaia Strengths (vs. Moonware) ​

  • Passive observation: does not require ground crew to carry devices or change workflows
  • Objective, automated timestamps: no human input required; impartial data
  • Prediction accuracy: POBT/PRDT based on actual visual progress of turnaround
  • Safety monitoring: CV can detect unsafe situations that GPS/telematics cannot (e.g., aerobridge not parked, FOD)
  • Emissions monitoring: APU detection via thermal/acoustic cameras
  • Scale of validation: 2M+ turnarounds, 21 airports, published benchmark reports
  • Deep airline relationships: IAG/BA equity investor; United, Alaska, Emirates deployments
  • A-CDM integration: directly feeds more accurate TOBT/POBT into A-CDM systems

Moonware HALO Strengths (vs. Assaia) ​

  • Active orchestration: dispatches crew and GSE to flights in real-time; on-demand task allocation
  • No camera infrastructure needed: eliminates cost/disruption of camera planning, installation, maintenance
  • Works beyond the gate: tracks GSE and crew across entire airfield, not just at stands
  • Autonomous vehicle readiness: designed as the control layer for future autonomous GSE
  • Faster deployment: software-only, 1-week implementation claimed
  • Lower upfront cost: no camera hardware or installation required
  • Ground traffic control: provides routing and coordination for ground vehicles
  • Spatial computing: HALO on Apple Vision Pro for future AR-based airfield management

Moonware's Explicit Argument Against Cameras ​

Moonware argues: "Camera-based systems are not only costly and time consuming to plan, install and maintain, but they can also be disruptive to airport operations during planning and implementation." Their position is that GPS/telematics + real-time data analytics can achieve coordination benefits without the infrastructure burden.

Complementary or Competitive? ​

These two companies serve overlapping but distinct functions:

  • Assaia answers: "What is happening at the gate right now, and when will the aircraft be ready?"
  • Moonware answers: "Which crew and equipment should go where, and how do they get there?"

In theory, they could be complementary -- Assaia's predictions feeding Moonware's dispatch engine. In practice, they compete for airport/airline budget and mindshare in the "ground operations AI" category. Assaia is far more mature (21 airports, $36M raised, 2M+ turns) versus Moonware's earlier-stage footprint.

11. Business Model ​

Revenue Model ​

  • SaaS subscription: recurring license fees per airport/per gate or per turnaround
  • Analysis as a Service: consulting/data analytics engagements
  • No hardware sales: Assaia uses existing cameras; does not sell/install camera infrastructure
  • Annual revenue: ~$4M (2025 estimate) -- still early revenue relative to deployment scale

Go-to-Market ​

  • Direct sales to airports and airlines
  • Strategic Partner Community (SPC): airports/airlines take equity stakes, provide product guidance, serve as reference customers
  • Exclusive distribution partnerships: Marubeni holds exclusive rights for Japan and Southeast Asia (since 2019/2023)
  • Technology partnerships: ADB SAFEGATE (VDGS cameras), SITA, Amadeus, INFORM, Copenhagen Optimization (gate allocation)

Customer Segments ​

  1. Airports -- primary buyer; deploys ApronAI, StandManager, SafetyControl, EmissionsControl
  2. Airlines -- deploys TurnaroundControl for their operations at specific airports
  3. Ground handlers -- uses TurnaroundControl and SafetyControl for SLA monitoring and safety compliance

12. Key Partnerships ​

PartnerTypeDetails
IAG (British Airways)Strategic investor + customerEquity stake; IAGi Accelerator origin; 116-gate Heathrow deployment
ADB SAFEGATETechnology integrationSafedock A-VDGS cameras feed ApronAI; joint go-to-market
MarubeniDistribution + investorExclusive ApronAI distribution for Japan & SE Asia; co-owner of Swissport Japan
GTAA (Toronto Pearson)Strategic investor + customerFounding SPC member; 106-gate full deployment
JetBlue VenturesStrategic investorFounding SPC member
ADR Ventures (Rome)Strategic investor + customerFirst continental European SPC member
Armira GrowthLead financial investorSeries B lead; operational support network
SITATechnology partnerAirport software integration
AmadeusTechnology partnerAODB and airport systems integration
INFORMTechnology partnerResource optimization integration
Copenhagen OptimizationTechnology partnerJoint intelligent gate allocation solution
Transformers GroupTechnology partnerCo-built StandManager (open-architecture RMS)
Emirates / Dubai AirportsCustomerFull-scale AI turnaround mgmt across DXB

13. Industry Recognition & Milestones ​

  • 2018: Gatwick becomes first airport worldwide to deploy CV on apron (Assaia)
  • 2019: British Airways becomes first airline to use AI video technology for punctuality (LHR T5)
  • 2022: Named TOP 100 Swiss Startups (Venturelab)
  • 2023: Strategic Partner Community launched with GTAA, IAG, JetBlue Ventures
  • 2024: Christiaan Hen (ex-Schiphol innovation chief) appointed CEO; millionth turnaround monitored
  • 2025: Copenhagen first airport with AI emissions monitoring; Munich and Dubai deployments; Series B ($26.6M); 2M+ turnarounds milestone; 21 airports
  • 2026: StandManager launched for gate optimization

14. Data Products & Reports ​

Turnaround Benchmark Report (Annual) ​

Published annually since 2023. The aviation industry's first standardized report on turnaround operations. Aggregates data from hundreds of thousands of AI-monitored turnarounds to benchmark performance across regions and airport sizes. Available as a gated download from assaia.com.

Key Data Assets Generated ​

  • Automated turnaround event timestamps (more accurate than manual/ACARS)
  • POBT/PRDT predictions
  • Safety incident and near-miss data
  • APU usage and emissions data
  • Stand utilization and buffer time analytics
  • Airline-by-airline and handler-by-handler performance benchmarks

15. Technology Risks & Limitations ​

  • Camera dependency: system cannot monitor events not visible to cameras (e.g., activities under the aircraft fuselage, inside the cabin). Moonware's GPS approach covers the full airfield.
  • Camera coverage gaps: not all stands at all airports have cameras with sufficient coverage; may require some camera additions/repositioning
  • Weather/lighting: outdoor cameras face occlusion from rain, snow, fog, darkness, sun glare. Assaia has not published robustness data for adverse conditions.
  • Privacy: video surveillance of ground crew raises labor/privacy concerns in some jurisdictions (especially EU under GDPR)
  • Revenue scale: ~$4M revenue against 21 airport deployments suggests many are still pilot/early-revenue stages
  • Model opacity: no published details on ML architectures, accuracy benchmarks per event type, false positive/negative rates, or model drift management

16. Key Takeaways for Competitive Analysis ​

  1. Assaia is the clear market leader in camera-based turnaround monitoring with 21 airports and 2M+ monitored turnarounds.

  2. The camera-first approach is both a strength and a constraint -- it provides objective, passive monitoring without requiring ground crew behavior change, but is limited to what cameras can see and requires camera infrastructure at every stand.

  3. Strategic investors (IAG, GTAA, Emirates/Dubai, ADR) provide deep distribution moats -- these relationships are hard for competitors to displace.

  4. Revenue is still small (~$4M) relative to deployment breadth, suggesting the business model is still scaling and many deployments may be pilot/proof-of-concept stages.

  5. Moonware and Assaia attack different parts of the problem. Assaia observes and predicts; Moonware orchestrates and dispatches. They are more complementary than directly competitive, but they compete for the same budget line items.

  6. The ADB SAFEGATE partnership is strategically important -- it gives Assaia access to thousands of VDGS-equipped stands globally without selling hardware.

  7. Prediction accuracy is the core value driver -- more accurate POBT/TOBT feeds directly into A-CDM systems, improving network-wide departure management. The AOBT study showed $4.5M/year savings from timestamp accuracy alone at a single US airport.

Sources ​

Public research notes collected from public sources.

Copyright © 2026 kvynlim