Deployment Playbook: Autonomous Vehicles at Airports

From First Test to Full Operations

Pre-Deployment Phase
Mapping and Route Planning
Vehicle Preparation
Testing Phases
Pilot Operations
Safety Validation
Scaling from Pilot to Operations
Airport Stakeholder Management
Training
Maintenance and Support
Operational Procedures
KPIs and Reporting
Multi-Airport Expansion
Insurance and Liability
Master Deployment Checklist

1. Pre-Deployment Phase

1.1 Site Survey

The site survey is the foundation of every airport deployment. It must capture the physical, operational, and regulatory characteristics of the airside environment before any vehicle arrives on site.

Physical Environment Assessment:

Measure all roadways, taxilanes, apron areas, and service roads with centimeter-level precision
Document surface types (asphalt, concrete, painted markings, expansion joints)
Record gradient changes, camber, drainage features, speed bumps, and bollards
Photograph and geolocate all signage, markings, lighting infrastructure, and physical barriers
Identify GPS shadow zones caused by terminal buildings, hangars, jet bridges, and other structures
Map Wi-Fi/cellular coverage across the operational area using signal strength surveys
Identify areas with electromagnetic interference (radar installations, ILS equipment, power lines)
Document weather exposure patterns: prevailing wind directions, flood-prone areas, ice accumulation zones

Operational Environment Assessment:

Map all vehicle traffic flows: GSE movements, fuel trucks, catering vehicles, de-icing units
Document aircraft push-back zones, taxi routes, and jet blast areas
Identify pedestrian crossing patterns and high-density personnel areas (gate areas, crew walkways)
Record peak traffic periods correlated to flight schedules
Identify FOD (Foreign Object Debris) hotspots and debris accumulation patterns
Document existing vehicle speed limits and traffic control measures
Map emergency vehicle access routes and staging areas

Infrastructure Assessment:

Locate available power supply for charging stations (voltage, amperage, proximity)
Identify suitable locations for edge compute infrastructure and communication base stations
Assess existing CCTV and surveillance coverage for integration opportunities
Document available facilities for vehicle staging, maintenance, and storage
Evaluate network infrastructure: fiber runs, switch locations, available bandwidth

1.2 Stakeholder Engagement

Internal Stakeholders (Airport Operator):

Airport CEO / Managing Director: Strategic alignment and board-level sponsorship
VP Operations / Airside Operations Director: Operational integration and safety culture owner
Safety Manager / SMS Coordinator: Safety case development and risk management
IT / Technology Director: Infrastructure, connectivity, and cybersecurity
Finance: Business case, capex/opex modeling, ROI tracking
Legal / Compliance: Contracts, liability, regulatory compliance
HR / Training: Workforce transition, operator recruitment, training programs
Communications / PR: Internal and external messaging

External Stakeholders:

FAA Regional Airport Certification Safety Inspector (Part 139 airports) or FAA Airports District Office (GA airports)
National aviation authority (CAA, EASA, CAAS, GCAA depending on jurisdiction)
Air Traffic Control (ATC) / Tower
Airlines operating at the airport
Ground handling companies
Fuel companies
Catering operators
Airport security / police
Local emergency services (fire, ambulance)
Insurance providers
Local government / planning authority
Neighboring communities (for landside operations)

Engagement Sequence:

Month 1-2: Internal alignment meeting with airport leadership. Build the business case. Identify executive sponsor.
Month 2-3: Informal engagement with FAA/CAA. Reference FAA CertAlert 24-02 for AGVS technology guidance. Discuss intended operational design domain (ODD).
Month 3-4: Formal notification to the FAA regional office or Airport Certification Safety Inspector. Present concept of operations (CONOPS).
Month 4-5: Stakeholder briefings with airlines, ground handlers, ATC. Collect concerns and requirements.
Month 5-6: Safety Management System (SMS) integration. Begin formal safety assessment.

1.3 Airport Authority Meetings

First Meeting (Concept Introduction):

Present the technology overview: vehicle capabilities, automation level (SAE Level 3-4), operational concept
Define the proposed operational design domain (ODD): routes, areas, times, weather limitations
Present relevant case studies (CVG, Changi, Teesside, Schiphol)
Discuss regulatory pathway and FAA/CAA guidance
Identify the approval process and decision-makers
Agree on communication cadence

Second Meeting (Safety and Integration):

Present preliminary hazard analysis
Discuss integration with existing Airport Emergency Plan (per 14 CFR 139.325)
Address ATC coordination requirements
Present proposed testing phases and timeline
Discuss insurance and liability framework
Review infrastructure requirements and costs

Third Meeting (Approval to Test):

Present completed Safety Risk Assessment (SRA)
Submit formal test plan with go/no-go criteria
Present the risk register with mitigations
Agree on monitoring and reporting requirements
Obtain written approval to proceed to testing phase

1.4 Safety Assessment

Hazard Identification Process: Conduct a structured hazard identification (HAZID) workshop with cross-functional participation including airport operations, ATC, ground handlers, AV engineering, and safety specialists.

Key Hazard Categories for Airport AV Operations:

Hazard Category	Example Hazards
Vehicle-Aircraft Interaction	Collision with taxiing aircraft, entering active taxi lane, jet blast exposure
Vehicle-Vehicle Interaction	Collision with manned GSE, fuel trucks, emergency vehicles
Vehicle-Person Interaction	Collision with ramp agents, passengers, maintenance crew
Infrastructure Damage	Collision with jet bridges, terminal buildings, lighting masts
FOD Generation	Vehicle component detachment, cargo spillage
Communication Failure	Loss of teleoperation link, loss of V2X connectivity
Sensor Degradation	Fog, heavy rain, glare, snow, sensor contamination
Cybersecurity	Remote vehicle compromise, GPS spoofing, data breach
Software Failure	Perception error, planning failure, localization drift
Operational	Incorrect route, unauthorized area entry, wrong cargo delivery

Safety Risk Assessment (SRA) Matrix:

Use the ICAO/FAA 5x5 risk matrix:

Severity: Catastrophic (A), Hazardous (B), Major (C), Minor (D), Negligible (E)
Likelihood: Frequent (5), Occasional (4), Remote (3), Improbable (2), Extremely Improbable (1)

All risks rated as "Unacceptable" (red zone) must be mitigated before operations begin. Risks rated "Tolerable" (yellow zone) require documented mitigation plans and ongoing monitoring.

1.5 Risk Register

The risk register is a living document that must be reviewed monthly during deployment and quarterly during steady-state operations.

Risk Register Structure:

Field	Description
Risk ID	Unique identifier (e.g., AV-R-001)
Hazard Description	Clear description of the hazard
Cause	Root cause or contributing factors
Consequence	Potential outcome if the hazard materializes
Pre-Mitigation Severity	A-E rating
Pre-Mitigation Likelihood	1-5 rating
Pre-Mitigation Risk Level	Calculated risk level
Mitigation Measures	Specific controls to reduce risk
Post-Mitigation Severity	A-E rating after controls
Post-Mitigation Likelihood	1-5 rating after controls
Post-Mitigation Risk Level	Residual risk level
Risk Owner	Named individual responsible
Review Date	Next scheduled review
Status	Open / Mitigated / Closed / Accepted

Sample Risk Register Entries:

Risk ID	Hazard	Pre-Mit Risk	Mitigation	Post-Mit Risk
AV-R-001	AV enters active taxiway	5A (Unacceptable)	Geofencing with redundant GNSS + INS, physical barriers, ATC coordination protocol	2C (Tolerable)
AV-R-002	AV collision with ramp agent	4B (Unacceptable)	LiDAR + camera pedestrian detection, 360-degree coverage, speed limiting in personnel zones, audible warnings	2D (Acceptable)
AV-R-003	Loss of communication link	4C (Unacceptable)	Dual-SIM cellular, dedicated Wi-Fi mesh, automatic safe-stop on comms loss, 30-second timeout	2D (Acceptable)
AV-R-004	Sensor degradation in heavy rain	4C (Unacceptable)	Rain-sensing algorithms (rated to 50mm/h per Changi trials), automatic speed reduction, safe-stop in extreme conditions	3D (Tolerable)
AV-R-005	Cybersecurity breach	3B (Unacceptable)	Network segmentation, encrypted V2X, OTA update authentication per UN R155/R156, vSOC monitoring	2D (Acceptable)
AV-R-006	Software perception error	4B (Unacceptable)	Redundant perception pipeline, sensor fusion validation, conservative planning margins, safety operator override	2C (Tolerable)

2. Mapping and Route Planning

2.1 How to Map an Airport for Autonomous Operations

Airport mapping for AV operations requires significantly higher accuracy than standard airport surveys. The goal is to create a centimeter-accurate digital twin of the operational environment.

Mapping Technology Stack:

Primary LiDAR: Vehicle-mounted 3D LiDAR (e.g., Velodyne Alpha Prime, Ouster OS1-128) for point cloud capture
RTK-GNSS: Real-Time Kinematic corrections achieving 1-4 cm horizontal accuracy (e.g., Point One Navigation, Trimble, u-blox)
IMU: High-grade Inertial Measurement Unit for dead-reckoning in GNSS-denied areas
Cameras: Calibrated stereo or multi-camera system for visual feature extraction and texture mapping
Odometry: Wheel encoders for ground-truth distance measurement

Mapping Process:

Step 1: Reference Network Establishment (1-2 days)

Establish ground control points (GCPs) across the operational area using survey-grade GNSS
Install or identify permanent reference markers for ongoing calibration
Validate RTK correction service coverage and accuracy across the site
Target accuracy: horizontal 2 cm, vertical 3 cm

Step 2: Data Collection Drives (1-3 days)

Drive the mapping vehicle along all intended routes and adjacent areas
Capture multiple passes at different times of day (varying lighting conditions)
Ensure coverage of all intersections, turning areas, parking positions, and charging locations
Collect data in both directions of travel
AeroVect has demonstrated digital twin creation of a major airport in under 2 hours using their Explorer mapping tool, though comprehensive mapping for operational deployment typically takes longer

Step 3: Point Cloud Processing (3-7 days)

Register and align point cloud data using SLAM (Simultaneous Localization and Mapping)
Apply RTK corrections for absolute positioning
Filter noise, remove dynamic objects (vehicles, aircraft, people)
Generate clean, geo-referenced 3D point cloud map

Step 4: Feature Extraction and Annotation (5-10 days)

Extract road boundaries, lane markings, stop lines, crosswalks
Identify and classify static objects: buildings, bollards, light poles, fences, signs
Annotate aircraft stand positions, gate numbers, and stand boundaries
Map speed zones, no-go areas, and transition zones
Create the three-layer HD map structure:
- Road Model: Road topology, elevation, surface type, gradient
- Lane Model: Lane boundaries, speed limits, directionality, permitted vehicle types
- Localization Model: Static features used for real-time localization (buildings, poles, curbs)

Step 5: Map Validation (2-3 days)

Drive validation routes with the AV stack in localization-only mode
Verify localization accuracy meets target (<10 cm lateral, <20 cm longitudinal)
Identify and resolve areas with poor localization (featureless zones, GPS shadows)
Document map accuracy metrics for the safety case

Step 6: Map Maintenance Plan

Schedule re-mapping after any construction or infrastructure changes
Implement incremental map updates from operational driving data
Define map version control and deployment procedures
Establish quarterly full-map validation drives

2.2 Waypoint Creation

Waypoint Types:

Route Waypoints: Define the driving path with 0.5-1.0 m spacing on straight segments, 0.2-0.5 m spacing on curves
Stop Waypoints: Precise positions for loading/unloading, charging, maintenance stops
Decision Waypoints: Points where the vehicle must evaluate conditions (intersections, merge points, aircraft stand entry)
Speed Transition Waypoints: Locations where speed limits change
Safety Waypoints: Emergency stop positions, safe harbor locations, pull-over zones

Waypoint Attributes: Each waypoint should include:

Latitude, longitude, altitude (WGS84)
Heading (desired vehicle orientation)
Maximum speed at this point
Allowed/required behaviors (stop, yield, proceed, wait for clearance)
Associated geofence zone ID
Lane assignment
Special conditions (e.g., "yield to aircraft," "check for jet blast," "pedestrian zone")

2.3 Geofencing

Geofencing is the primary safety boundary mechanism for airport AV operations. It operates at multiple layers.

Geofence Layer Structure:

Layer	Purpose	Action on Breach
Hard Boundary	Absolute no-go zone (runways, taxiways, restricted areas)	Immediate emergency stop
Soft Boundary	Operational limit (edge of permitted route with buffer)	Speed reduction + alert to operator
Speed Zone	Area-specific speed limit	Automatic speed enforcement
Operational Zone	Permitted operational area	Normal operations within zone
Dynamic Zone	Temporary restriction (aircraft on stand, maintenance area)	Route recalculation or hold

Geofence Implementation Requirements:

Redundant positioning: GNSS + INS + visual localization. No single-source dependency.
Geofence check rate: minimum 10 Hz (every 100 ms)
Position uncertainty must be factored into geofence margins (if position uncertainty is 0.5 m, geofence buffer must be >0.5 m from the actual boundary)
Geofence violations must be logged, time-stamped, and reported
Dynamic geofence updates must propagate to all vehicles within 5 seconds

Critical Geofence Zones for Airports:

Runway and taxiway boundaries (absolute no-go)
Active aircraft stand zones (dynamic, based on flight schedule)
Jet blast exclusion zones (dynamic, based on aircraft engine status)
Fuel farm perimeter
ILS critical and sensitive areas
Construction zones (dynamic)
Emergency vehicle staging areas

2.4 NOTAM Integration

NOTAMs (Notices to Air Missions) contain real-time operational status information that directly affects AV route planning. Legacy NOTAM systems are text-based and not well-suited for automated parsing, but modern APIs enable integration.

NOTAM Integration Architecture:

Data Source: Connect to FAA NOTAM Search API or commercial NOTAM parsing service (e.g., Notamify API with Atomic Elements for structured data)
Parsing Engine: Automated classification of NOTAMs relevant to ground operations:
- Runway/taxiway closures affecting crossing routes
- Construction activities on or near AV routes
- Temporary obstacles or hazards
- Changed radio frequencies or communication procedures
- Airfield lighting outages
Route Impact Assessment: Automated evaluation of whether active NOTAMs affect any planned AV routes
Dynamic Route Adjustment: Automatic rerouting or operational hold when NOTAM impacts are detected
Operator Notification: Alert to human supervisor for NOTAMs requiring operational decisions

Integration with Flight Data:

Connect to Airport Collaborative Decision Making (A-CDM) system for real-time flight status
Correlate aircraft arrival/departure times with AV routing to avoid conflicts
Integrate with gate management system for stand allocation awareness
Use flight schedule data to predict peak traffic periods and adjust AV operations accordingly

3. Vehicle Preparation

3.1 Sensor Mounting

Sensor Suite for Airport Airside Operations:

Sensor	Quantity	Placement	Purpose
3D LiDAR (long range)	1-2	Roof center/front	Primary perception, 200+ m range
3D LiDAR (short range)	2-4	Corners, low mount	Close-range 360-degree coverage, blind spot elimination
Camera (forward)	1-2	Windshield area	Forward perception, traffic light/sign recognition
Camera (surround)	4-8	Distributed around vehicle	360-degree visual perception
Radar (forward)	1-2	Front bumper	Long-range velocity detection, weather resilience
Radar (rear/side)	2-4	Bumper/fender	Side/rear object detection
Ultrasonic	8-12	Bumper perimeter	Very close-range parking/docking
GNSS antenna	2	Roof (dual antenna)	Position + heading from RTK-GNSS
IMU	1	Vehicle center of mass	Dead reckoning, orientation
Thermal camera	1-2	Front/rear	Jet blast detection, personnel detection in low visibility
V2X antenna	1-2	Roof	Vehicle-to-everything communication

Mounting Standards:

All sensor mounts must be vibration-isolated and thermally managed
Mounting brackets must withstand airport operating conditions: jet blast (up to 35 mph sustained), rain, temperature extremes (-20C to +50C)
Sensor positions must not increase vehicle height beyond airport clearance limits
Mounting must allow field replacement of individual sensors without full recalibration
All electrical connections must be waterproof (IP67 minimum for external sensors)
Cable routing must avoid pinch points, heat sources, and moving parts

3.2 Calibration

Intrinsic Calibration (per sensor):

Camera: lens distortion parameters, focal length, principal point
LiDAR: beam angle offsets, range calibration, intensity calibration
Radar: beam pattern verification, range/velocity accuracy
IMU: bias estimation, scale factor, cross-axis sensitivity
Frequency: At installation, after any sensor replacement, and monthly verification

Extrinsic Calibration (sensor-to-sensor and sensor-to-vehicle):

LiDAR-to-camera: 6-DOF transformation using calibration targets. Target accuracy: <1 degree rotation, <2 cm translation
LiDAR-to-LiDAR: Point cloud registration between multiple LiDAR units
LiDAR-to-vehicle: Alignment to vehicle coordinate frame using ground plane and known reference points
Camera-to-camera: Multi-camera rig calibration for consistent surround view
GNSS antenna-to-IMU: Lever arm measurement with survey-grade accuracy
Frequency: At installation, after any physical impact, and monthly verification

Calibration Validation:

Drive a calibration route with known reference points
Verify cross-sensor object position agreement (<5 cm at 30 m range)
Verify LiDAR-camera projection accuracy (❤️ pixels at 50 m)
Automated calibration health monitoring during operation
Alert and safe-stop if calibration drift exceeds thresholds

3.3 Compute Installation

Compute Hardware Requirements:

Component	Specification	Example
Primary compute	GPU-accelerated AI inference, 200+ TOPS	NVIDIA DRIVE AGX Orin/Thor
Secondary compute	Safety controller, redundant path	ARM-based safety MCU
Storage	High-speed SSD for data logging	2-4 TB NVMe
Network switch	In-vehicle Ethernet backbone	Automotive Ethernet (1000BASE-T1)
Power supply	Regulated DC-DC converter	12V/24V to compute voltages

Installation Requirements:

Compute unit must be mounted in a sealed, climate-controlled enclosure
Operating temperature range: -20C to +70C (with active cooling)
Vibration rating: MIL-STD-810G or equivalent
IP67 enclosure rating for external mounting, IP54 minimum for in-cabin
Dedicated power circuit with battery backup (minimum 60 seconds for safe shutdown)
EMI shielding to prevent interference with aircraft navigation systems (DO-160G or equivalent testing)
Heat dissipation: ensure compute does not overheat in direct sunlight on hot tarmac (up to 60C+ surface temperatures)

Functional Safety Architecture:

Primary compute runs perception, planning, and control
Independent safety controller monitors primary compute outputs and can override
Watchdog timer: if primary compute fails to send heartbeat within 100 ms, safety controller triggers safe stop
Dual-redundant power paths to compute system
Compliant with ISO 26262 ASIL-D for safety-critical functions

3.4 Communication Setup

Communication Stack:

Link	Technology	Purpose	Latency Target
Primary cellular	5G / LTE	Teleoperation, fleet management, OTA updates	<100 ms
Secondary cellular	Different carrier (dual-SIM)	Redundant connectivity	<100 ms
Wi-Fi mesh	802.11ax (Wi-Fi 6)	High-bandwidth data offload, local connectivity	<20 ms
V2X	C-V2X (PC5 sidelink)	Direct vehicle-to-vehicle and vehicle-to-infrastructure	<20 ms
DSRC	802.11p (5.9 GHz)	Legacy V2X compatibility if required	<20 ms
Satellite	L-band backup	Emergency position reporting	<2000 ms

Communication Requirements:

Minimum 10 Mbps uplink for 4K video streaming to remote operations center
Minimum 2 Mbps downlink for commands and map updates
Maximum acceptable round-trip latency for teleoperation: 150 ms (100 ms target)
Automatic failover between communication paths within 2 seconds
If all communication is lost, vehicle must execute safe-stop procedure within 5 seconds
Communication encryption: TLS 1.3 minimum for all links
Communication link quality monitoring at 1 Hz with alerts on degradation

Remote Operations Center (ROC) Requirements:

Dedicated monitoring stations with multi-screen displays (minimum 3 screens per operator)
Real-time video feeds from all active vehicles
Fleet management dashboard showing vehicle positions, status, and alerts
One-click emergency stop capability for any vehicle
Voice communication with field personnel
Secure, redundant network connectivity to ROC
UPS backup power for ROC (minimum 4 hours)
Operator-to-vehicle ratio: start at 1:1 during pilot, target 1:3 to 1:5 at scale

3.5 Safety Systems Validation

Emergency Stop (E-Stop) Systems:

E-Stop Type	Trigger	Response Time	Action
Hardware E-Stop button	Manual press by safety operator	<100 ms	Immediate brake application, power to drive motors cut
Software E-Stop	Autonomous detection of unsafe condition	<200 ms	Controlled deceleration to stop
Remote E-Stop	Operator command from ROC	<500 ms (includes comms latency)	Controlled deceleration to stop
Geofence E-Stop	Hard geofence boundary breach	<200 ms	Immediate brake application
Watchdog E-Stop	Compute heartbeat timeout (100 ms)	<300 ms	Controlled deceleration to stop
Communication loss E-Stop	All comms lost for >5 seconds	<200 ms after timeout	Controlled deceleration to stop

Validation Testing Requirements:

Test each E-Stop mechanism independently: minimum 10 activations each
Test at maximum operating speed
Measure and record actual response times
Verify braking distance at maximum speed and maximum payload
Test E-Stop under degraded conditions (wet surface, slope, partial brake failure)
Verify vehicle is safe and stable after E-Stop (no rollaway, hazard lights activated)
Document all results in the safety case

4. Testing Phases

4.1 Factory Acceptance Test (FAT)

The FAT is conducted at the vehicle manufacturer's or integrator's facility before the vehicle ships to the airport. It validates that the vehicle meets specifications and is ready for site testing.

FAT Checklist:

Mechanical Systems:

[ ] Vehicle dimensions and weight within specification
[ ] Steering system: full lock-to-lock operation, no binding
[ ] Brake system: service brake and parking brake functional
[ ] Suspension: no abnormal noises or leaks
[ ] Tires: correct specification, pressure, condition
[ ] Lighting: headlights, taillights, turn signals, hazard lights, beacon
[ ] Horn / audible warning system functional
[ ] Towing attachment / cargo interface functional
[ ] All body panels secure, no sharp edges

Electrical Systems:

[ ] Battery capacity test: meets range specification (with margin)
[ ] Charging system: compatible with planned charger, charges to full
[ ] 12V/24V auxiliary systems functional
[ ] All fuses and circuit breakers correctly rated
[ ] Emergency power-off functional
[ ] Wiring harness: no chafing, all connectors secure

Sensor Systems:

[ ] All sensors powered and returning data
[ ] Intrinsic calibration verified for each sensor
[ ] Extrinsic calibration verified (sensor-to-sensor alignment)
[ ] LiDAR: point cloud clean, no blind spots in coverage
[ ] Cameras: image quality, no lens defects, correct exposure
[ ] Radar: target detection verified at specified ranges
[ ] GNSS: position fix acquired, RTK corrections working
[ ] IMU: drift rates within specification

Compute and Software:

[ ] Compute system boots successfully, all processes start
[ ] Perception pipeline: detects objects in test environment
[ ] Planning module: generates valid trajectories
[ ] Control module: vehicle follows trajectories smoothly
[ ] Localization: converges on HD map in test area
[ ] Data logging: all sensor data recorded correctly
[ ] Software version matches release specification

Communication:

[ ] Cellular connectivity: both SIMs connect
[ ] Wi-Fi: connects to test network
[ ] V2X: transmits and receives messages
[ ] Remote operations interface: video streaming, command reception
[ ] Remote E-Stop: functional via ROC

Safety Systems:

[ ] Hardware E-Stop: tested from all button locations
[ ] Software E-Stop: triggered by test conditions
[ ] Geofence: vehicle stops at test geofence boundary
[ ] Watchdog: triggers on simulated compute failure
[ ] Communication loss: safe-stop on simulated comms failure
[ ] Collision avoidance: AEB triggers on test targets

Documentation:

[ ] Vehicle build record complete
[ ] Sensor serial numbers and calibration certificates
[ ] Software version and configuration record
[ ] Test results documented and signed
[ ] Known issues list (if any) with severity ratings
[ ] Spare parts list provided
[ ] Maintenance manual provided

4.2 Site Acceptance Test (SAT)

The SAT is conducted at the airport after vehicle delivery. It validates that the vehicle operates correctly in the actual deployment environment.

SAT Pre-Conditions:

Airport authority approval for testing obtained
Test area secured and access controlled
Safety operator trained and certified
ROC set up and communication verified
HD map of test area loaded and validated
Emergency response plan in place
Weather conditions within operational limits

SAT Test Sequence:

Phase 1: Static Tests (Day 1)

[ ] GNSS position accuracy at 10+ locations across test area (<5 cm with RTK)
[ ] Cellular signal strength survey across test area
[ ] Sensor health checks at multiple positions
[ ] HD map localization convergence test
[ ] Communication link quality to ROC verified
[ ] E-Stop tests at the airport site (all types)

Phase 2: Low-Speed Controlled Driving (Days 2-3)

[ ] Manual driving through all planned routes to verify drivability
[ ] Autonomous driving on straight segments at 5 km/h
[ ] Autonomous driving through curves at 5 km/h
[ ] Stop-and-go at designated waypoints
[ ] Obstacle detection: static object placed in path
[ ] Pedestrian detection: person walking in front of vehicle
[ ] Geofence compliance: vehicle approaches boundary and stops

Phase 3: Operational Speed Testing (Days 4-5)

[ ] Autonomous driving at planned operational speed (typically 15-25 km/h for airside)
[ ] Lane keeping accuracy on straight and curved sections
[ ] Intersection handling: yield, stop, proceed
[ ] Dynamic obstacle avoidance: vehicle or person entering path
[ ] Multi-vehicle interaction (if multiple vehicles available)
[ ] Night driving test (if night operations planned)
[ ] Cargo loading/unloading docking accuracy

Phase 4: Stress Testing (Days 6-7)

[ ] Continuous operation for 4+ hours
[ ] Operation during different weather conditions (if available during test window)
[ ] Repeated route execution: 20+ consecutive runs
[ ] Edge case testing: tight turns, narrow passages, uneven surfaces
[ ] Communication degradation testing: measure performance with reduced signal
[ ] Recovery from safe-stop: vehicle resumes operations correctly

4.3 Integration Testing

Integration testing validates that the AV system works correctly within the broader airport operational ecosystem.

Integration Test Areas:

Test Category	Test Items
Fleet Management	Vehicle assignment, route dispatch, status monitoring, multi-vehicle coordination
A-CDM Integration	Flight data reception, schedule-based route timing, gate allocation awareness
NOTAM Integration	NOTAM parsing, route impact assessment, dynamic rerouting
Airport Systems	FIDS integration, gate management, resource planning
Ground Handler Systems	Task assignment, cargo tracking, turnaround coordination
Communication	ROC-to-vehicle, vehicle-to-vehicle, vehicle-to-infrastructure
Data Pipeline	Operational data collection, cloud upload, analytics dashboard
Charging Infrastructure	Automated charging initiation, charge management, scheduling

4.4 Operational Readiness Review (ORR)

The ORR is a formal gate review before transitioning from testing to pilot operations. It requires sign-off from all key stakeholders.

ORR Go/No-Go Criteria:

Category	Criteria	Status
Safety	All unacceptable risks mitigated to tolerable or acceptable	[ ] Go / [ ] No-Go
Safety	Safety case document complete and reviewed	[ ] Go / [ ] No-Go
Safety	Emergency response plan tested and approved	[ ] Go / [ ] No-Go
Technical	All FAT items passed	[ ] Go / [ ] No-Go
Technical	All SAT items passed	[ ] Go / [ ] No-Go
Technical	Integration testing complete, no critical issues open	[ ] Go / [ ] No-Go
Technical	HD map validated, localization accuracy confirmed	[ ] Go / [ ] No-Go
Operational	Safety operators trained and certified	[ ] Go / [ ] No-Go
Operational	ROC staffed and operational procedures documented	[ ] Go / [ ] No-Go
Operational	Maintenance plan in place, spares available	[ ] Go / [ ] No-Go
Regulatory	Airport authority written approval for pilot operations	[ ] Go / [ ] No-Go
Regulatory	FAA/CAA notification complete (if required)	[ ] Go / [ ] No-Go
Regulatory	Insurance coverage active and sufficient	[ ] Go / [ ] No-Go
Stakeholder	Airlines and ground handlers briefed and acknowledged	[ ] Go / [ ] No-Go
Stakeholder	ATC coordination protocol agreed	[ ] Go / [ ] No-Go
Support	Incident reporting system in place	[ ] Go / [ ] No-Go
Support	Escalation procedures documented	[ ] Go / [ ] No-Go

Decision: All criteria must be "Go" to proceed. Any "No-Go" requires documented remediation plan with timeline.

5. Pilot Operations

5.1 How to Run a Pilot (1-3 Vehicles)

Pilot Scope Definition:

Parameter	Recommended Pilot Scope
Fleet size	1-3 vehicles
Duration	6-12 months
Operating hours	Defined shift (e.g., 06:00-22:00), expanding over time
Routes	1-3 fixed routes, well-characterized
Speed	Conservative (10-20 km/h, below site maximum)
Weather limits	Clear to moderate conditions initially
Cargo/payload	Non-critical loads initially (training baggage, non-time-sensitive cargo)
Automation level	SAE Level 3-4 with human safety operator on or near vehicle

Phased Pilot Approach:

Phase A: Shadow Mode (Weeks 1-4)

Vehicle operates with a safety driver who controls the vehicle manually
Autonomous system runs in parallel, making decisions but not controlling the vehicle
Compare autonomous decisions to human driver actions
Identify edge cases and system limitations
Target: 200+ operational hours in shadow mode

Phase B: Supervised Autonomy (Weeks 5-16)

Vehicle operates autonomously with safety operator in the vehicle
Safety operator can intervene at any time via steering wheel, brake, or E-Stop
Document all interventions (disengagements) with detailed root cause analysis
Target: Demonstrate improving disengagement rate (industry benchmark trend: ~16.7% reduction per year of testing)
Initial target: <1 disengagement per 10 km, improving to <1 per 50 km

Phase C: Remote Supervision (Weeks 17-26)

Safety operator moves from vehicle to a follow vehicle or fixed observation point
Remote operator monitors via ROC with intervention capability
Safety operator remains within line-of-sight initially, then moves to ROC-only
Requires demonstrated reliability: <1 disengagement per 100 km
Airport authority approval required for this transition

Phase D: Routine Operations (Weeks 27-52)

Vehicle operates with ROC monitoring only (no on-site safety operator per vehicle)
1 safety operator available as roving support for 1-3 vehicles
Formal performance reporting to airport authority (monthly)
Continuous improvement based on data analysis

5.2 Scope Limitations

During pilot operations, the following limitations should be clearly defined and communicated:

No mixed traffic with aircraft under power: Initially, do not operate in areas where aircraft are actively taxiing under their own power
No runway crossings: Even if the route conceptually requires it, avoid runway crossings during pilot phase
Weather restrictions: Define clear weather minimums (e.g., visibility >500 m, wind <40 km/h, no active thunderstorm, rain intensity <50 mm/h)
Time restrictions: Avoid peak traffic periods initially; operate during off-peak hours
Cargo restrictions: No dangerous goods, no live animals, no high-value or time-critical cargo
Speed restrictions: Maximum 15-20 km/h during pilot, regardless of route speed limits
Passenger restrictions: No passengers in/on the vehicle during initial pilot

5.3 Human Supervision Requirements

FAA Guidance (per CertAlert 24-02 and Bulletin 25-02):

A human operator must remain capable of regaining instantaneous control of the AGVS should the automated system fail
When operating around moving aircraft, airport employees, vehicles, and equipment, a human monitor should be physically located in/near the AGVS
The human monitor must be properly badged/escorted by the airport and trained in airport policies

Supervision Ratio Progression:

Phase	Ratio (Operator:Vehicle)	Operator Location
Shadow Mode	1:1	In vehicle
Supervised Autonomy	1:1	In vehicle
Remote Supervision (early)	1:1	Follow vehicle / observation point
Remote Supervision (mature)	1:2	ROC + roving field support
Routine Operations	1:3 to 1:5	ROC + roving field support
Full Scale	1:5 to 1:10	ROC only + on-call field support

5.4 Data Collection

Operational Data (Collected Continuously):

Vehicle telemetry: position, speed, heading, steering angle, brake pressure, motor torque
Sensor data: raw LiDAR, camera, radar data (stored for replay and analysis)
Perception outputs: detected objects, classifications, tracked trajectories
Planning outputs: planned trajectories, decision points, yield/go decisions
Control outputs: steering commands, throttle/brake commands, actual vs. planned deviation
Communication metrics: signal strength, latency, packet loss, failover events
System health: compute temperature, memory usage, sensor status, battery state

Safety-Critical Events (Flagged and Reviewed Within 24 Hours):

All disengagements (human takeover) with categorization:
- Safety disengagement: intervention to prevent unsafe condition
- Non-safety disengagement: intervention for comfort, efficiency, or operational reasons
- System-initiated disengagement: vehicle requested human takeover
All E-Stop activations
All near-miss events (object detected within safety margin)
All geofence alerts
All communication loss events
All perception anomalies (false positive/negative detections)

Performance Metrics (Calculated Daily):

Distance driven (autonomous vs. manual)
Hours operated (autonomous vs. manual)
Missions completed vs. assigned
Disengagement rate per km and per hour
Average speed (moving vs. overall)
On-time performance for scheduled missions
Energy consumption per km

6. Safety Validation

6.1 How Many Hours/Km for Safety Case

The RAND Corporation's landmark study "Driving to Safety" (Kalra & Paddock, 2016) established that demonstrating AV safety purely through miles driven is statistically impractical for rare events like fatalities, requiring hundreds of billions of miles. However, airport operations differ fundamentally from public roads:

Airport-Specific Safety Case Approach:

Rather than attempting to demonstrate statistical safety through mileage alone, airport AV safety cases should use a multi-pillar approach:

Pillar	Description	Minimum Threshold
Simulation Testing	Virtual testing of perception, planning, and control	100,000+ simulated scenarios covering all identified hazards
Closed-Course Testing	Track testing of safety-critical scenarios	500+ structured test cases, all passed
Operational Testing	Real-world driving in the airport environment	5,000+ autonomous km with <1 safety disengagement per 500 km
Safety Audit	Independent review of safety processes, design, and operations	Complete audit with no critical findings
In-Service Monitoring	Continuous monitoring during operations	Ongoing, with defined escalation triggers

Recommended Minimum Operational Testing Before Moving to Unsupervised Operations:

10,000 autonomous km without a safety-critical disengagement
2,000 operational hours without a safety-critical incident
500+ loading/unloading cycles without a safety event
50+ successful adverse weather operations (rain, low visibility, wind)
Zero collisions with aircraft, vehicles, infrastructure, or personnel
Demonstrated improvement trend in disengagement rate over time

Note: These thresholds are guidelines based on industry practice. The actual requirements will be determined through negotiation with the airport authority and aviation regulator, informed by the specific risk profile of the operational design domain.

6.2 Scenario Testing Requirements

Scenario Categories and Example Tests:

Category 1: Normal Operations

Drive full route from start to destination at operational speed
Navigate all intersections and yield points
Complete loading/unloading at designated positions
Return to charging station and initiate charging
Operate for full shift duration

Category 2: Dynamic Object Interaction

Pedestrian crossing path at various distances and speeds
Vehicle crossing path at intersection
Vehicle approaching from behind (overtaking scenario)
Stationary obstacle in lane (parked vehicle, equipment, FOD)
Multiple simultaneous objects requiring priority decision
Aircraft pushback across AV route

Category 3: Degraded Conditions

Rain (light, moderate, heavy up to 50 mm/h)
Fog (visibility 200 m, 500 m, 1000 m)
Night operations (with and without apron lighting)
Wet surface (reduced traction)
Glare (sun angle, reflections from aircraft)
Wind gusts (30-50 km/h)
Single sensor failure (LiDAR, camera, radar, GNSS)
Single communication link failure
Partial compute degradation

Category 4: Emergency Scenarios

E-Stop from all trigger sources
Loss of all communication
Complete GNSS failure
Multiple simultaneous sensor failures
Vehicle mechanical failure (tire, steering, brake warning)
Fire or smoke detection on vehicle
Airport emergency (aircraft emergency, security incident)
Power failure at charging station

Category 5: Edge Cases (Airport-Specific)

Jet blast encounter (thermal and aerodynamic detection)
FOD on driving surface
Water pooling on apron
Ice or snow on driving surface
Construction zone with changed layout
Unusual vehicle (oversized load, emergency vehicle with lights)
Aircraft under tow crossing path
Personnel running (emergency evacuation scenario)
Flocks of birds on apron

6.3 Incident Response Plan

Incident Classification:

Level	Description	Response Time	Notification
Level 1: Critical	Collision with aircraft, vehicle, person, or infrastructure causing injury or significant damage	Immediate	Airport Ops, Emergency Services, ATC, AV Company CEO, Regulator (within 2 hours)
Level 2: Serious	Near-miss with aircraft or person, property damage without injury, vehicle fire	<15 minutes	Airport Ops, AV Operations Manager, Safety Manager (within 4 hours)
Level 3: Moderate	Vehicle stops unexpectedly blocking operations, cargo damage, minor vehicle damage	<30 minutes	AV Operations Manager, Airport Ops (within 24 hours)
Level 4: Minor	System anomaly, unexpected disengagement, communication loss recovered	<2 hours	AV Operations Team (within 24 hours)

Incident Response Procedure:

Immediate Actions (first 5 minutes):

Activate E-Stop if not already triggered
Ensure safety of all personnel in the area
Notify Airport Operations Control (radio channel designated for AV operations)
If injury: call emergency services, begin first aid
If blocking aircraft operations: notify ATC immediately
Secure the scene, prevent further damage
Activate vehicle hazard lights and deploy warning cones

Short-Term Actions (5-60 minutes):

Deploy incident response team to scene
Document the scene: photographs, measurements, witness statements
Preserve all vehicle data logs (do not power cycle unless necessary for safety)
If vehicle can be moved safely: relocate to non-operational area
If vehicle cannot be moved: arrange towing and notify affected operations
Begin preliminary incident report
Notify insurance provider (for Level 1 and 2 incidents)

Investigation Phase (1-7 days):

Download and preserve all sensor data, logs, and telemetry
Reconstruct the incident using data replay tools
Identify root cause and contributing factors
Determine if the incident reveals a systemic issue
If systemic: suspend operations pending resolution
If isolated: document corrective actions and resume with enhanced monitoring

Post-Incident Actions:

Complete formal incident report
Update risk register with new or revised risk entries
Implement corrective actions
Conduct lessons-learned session with all stakeholders
Submit regulatory notifications as required
Update training materials if needed
Resume operations with formal approval from Safety Manager

Operational Pause Triggers (Suspend All AV Operations Immediately):

Any collision with an aircraft
Any injury to a person caused by the AV
Multiple Level 2 incidents within 7 days
Systemic software or hardware failure affecting safety functions
Regulator direction to cease operations
Airport authority direction to cease operations

7. Scaling from Pilot to Operations

7.1 Adding Vehicles

Fleet Expansion Criteria (Prerequisites Before Adding Each Batch):

Current fleet achieving target KPIs for 30 consecutive days
No open Level 1 or Level 2 incidents
Disengagement rate below target threshold
ROC capacity and operator staffing sufficient for expanded fleet
Charging infrastructure sufficient for expanded fleet
Maintenance capacity (tools, spares, trained technicians) scaled
Insurance coverage updated for expanded fleet
Airport authority approval for expansion

Recommended Expansion Schedule:

Phase	Fleet Size	Duration	Key Milestone
Pilot Phase A	1 vehicle	Months 1-3	Complete SAT, begin shadow mode
Pilot Phase B	1-2 vehicles	Months 4-6	Supervised autonomy, first disengagement targets met
Pilot Phase C	2-3 vehicles	Months 7-12	Remote supervision, routine operations demonstrated
Early Operations	3-5 vehicles	Months 13-18	Multiple routes, off-peak hours
Growing Operations	5-10 vehicles	Months 19-24	Extended hours, peak period introduction
Full Operations	10-20+ vehicles	Months 25+	24/7 capability, full route network

7.2 Expanding Routes

Route Expansion Process:

Survey: Map new route area using established mapping process
Risk Assessment: Conduct SRA for new route, update risk register
Geofencing: Define and validate geofences for new area
Testing: Run SAT test sequence on new route (abbreviated version, 2-3 days)
Shadow Mode: Operate in shadow mode on new route (1-2 weeks)
Supervised: Transition to supervised autonomy on new route
Integration: Add new route to fleet management system and operational schedule
Approval: Obtain airport authority sign-off for new route
Operational: Begin routine operations on new route

7.3 Reducing Human Supervision

Supervision Reduction Decision Framework:

Metric	Threshold for Supervision Reduction
Autonomous km since last safety disengagement	>2,000 km
Operational hours since last safety disengagement	>500 hours
Total non-safety disengagement rate	<1 per 200 km
E-Stop activations	Zero in last 30 days
Near-miss events	<1 per 1,000 km
System availability (uptime)	>95%
Communication link availability	>99.5%
Airport authority comfort level	Documented approval

Supervision reduction must be approved by:

AV Safety Manager
Airport Operations Manager
Airport Safety Manager
FAA/CAA (if required by terms of approval)

7.4 Fleet Management

Fleet Management System Requirements:

Real-time vehicle tracking: Position, speed, heading, status for all vehicles on a single dashboard
Mission management: Automated task assignment based on demand, vehicle availability, and route optimization
Charging management: Automated charge scheduling to ensure vehicle availability while minimizing energy costs
Health monitoring: Continuous vehicle health tracking with predictive maintenance alerts
Data management: Centralized collection, storage, and analysis of all operational data
Reporting: Automated daily, weekly, and monthly performance reports
Alerting: Tiered alert system with escalation to appropriate personnel
Map management: Version-controlled HD map deployment to fleet
Software management: Staged OTA software deployment with rollback capability
Compliance: Audit trail for all operational decisions and system changes

Fleet Operations Center Staffing:

Shift	Fleet Size 1-3	Fleet Size 4-10	Fleet Size 11-20	Fleet Size 20+
Day (peak)	1 operator	2 operators + 1 supervisor	3 operators + 1 supervisor	4+ operators + 1 supervisor
Evening	1 operator	1 operator	2 operators + 1 supervisor	3+ operators + 1 supervisor
Night (if 24/7)	1 operator	1 operator	1 operator + 1 on-call	2 operators + 1 on-call
Field support	1 roving	1 roving	2 roving	1 per 10 vehicles

8. Airport Stakeholder Management

8.1 Who Needs to Approve What

Approval Matrix:

Stakeholder	What They Approve	When Needed	How to Engage
Airport Authority / Operator	Overall permission to operate on airport property; specific areas, routes, and times; access to airside	Before any on-airport activity	Formal meetings, written agreements, regular progress reports
FAA (Part 139 airports)	Review and awareness of AGVS operations; may require modifications to Airport Certification Manual (ACM)	Before testing in movement areas; early notification for non-movement areas	Contact Regional Airport Certification Safety Inspector per CertAlert 24-02
National CAA (non-US)	Regulatory approval depending on jurisdiction; may require formal safety case submission	Varies by country; engage early	Formal application process; reference ICAO standards
ATC / Tower	Coordination procedures for areas near movement areas; communication protocols	Before any operations near taxiways or runways	Joint meetings, agreed communication protocol document
Airlines	Acknowledgment of AV operations; integration with turnaround process; data sharing for A-CDM	Before pilot operations on routes serving their gates	Airline station manager briefings; written notification
Ground Handling Companies	Integration with ground handling workflows; deconfliction of movements; personnel safety briefing	Before pilot operations on shared apron areas	Joint operational procedures; training for GH staff
Fuel Companies	Deconfliction of fuel truck movements; safety in fuel zones	Before operations near fuel hydrant systems or fuel farms	Safety briefing; operating procedure agreement
Airport Security	Vehicle access credentials; background checks for operators; cybersecurity review	Before first vehicle enters airside	Security department engagement; compliance documentation
Airport Fire Service	Emergency response procedures for AV incidents; vehicle familiarization	Before pilot operations	Fire service briefing; familiarization session with AV
Insurance Provider	Coverage for AV operations; liability terms	Before first operational movement	Policy negotiation; risk assessment submission
Local Regulator (vehicles)	Vehicle registration or exemption (if operating on public roads to access airport)	If AV transits public roads	Application per local AV regulations

8.2 Stakeholder Communication Plan

Regular Communications:

Communication	Audience	Frequency	Content
Progress Report	Airport Authority, FAA/CAA	Monthly	KPIs, incidents, upcoming milestones
Operations Brief	Airlines, Ground Handlers	Bi-weekly during pilot, monthly at scale	Schedule changes, route expansions, lessons learned
Safety Report	Airport Safety Committee	Quarterly	Safety metrics, risk register updates, incident analysis
Executive Update	Airport CEO, AV Company leadership	Quarterly	Strategic progress, business case tracking, expansion plans
ATC Coordination	ATC / Tower	As needed, minimum monthly	Route changes, schedule changes, communication protocol updates
Awareness Notice	All airport tenants and personnel	Before each major phase change	What to expect, safety information, contact details

9. Training

9.1 Operator Training

Safety Operator / Vehicle Operator Training Program:

Module	Duration	Content
Module 1: Airport Fundamentals	8 hours	Airside safety, ramp rules, FOD awareness, emergency procedures, radio communications, badge requirements
Module 2: Vehicle Systems	16 hours	AV architecture overview, sensor systems, compute system, communication systems, E-Stop systems, manual driving controls
Module 3: Autonomous Operations	16 hours	How the AV perceives, plans, and acts; understanding system limitations; recognizing degraded performance; ODD boundaries
Module 4: Intervention and Takeover	16 hours	When and how to intervene; disengagement procedures; manual override techniques; practice scenarios on closed course
Module 5: Normal Operations	8 hours	Pre-trip checks, startup/shutdown, mission management, charging procedures, shift handover, data reporting
Module 6: Emergency Procedures	8 hours	E-Stop activation, incident response, vehicle recovery, emergency communication, airport emergency plan integration
Module 7: ROC Operations	8 hours	Remote monitoring interface, teleoperation (if applicable), fleet management dashboard, alert handling, escalation
Module 8: Practical Assessment	16 hours	Supervised operation of AV on actual routes; emergency scenario exercises; written and practical examination

Total Training Duration: 96 hours (approximately 12 days)

Certification Requirements:

Pass written examination (minimum 80% score)
Pass practical assessment (assessed by qualified examiner)
Complete 40 hours of supervised operational experience
Recertification annually with 8-hour refresher course
Additional training required when new routes, vehicles, or software versions are introduced

9.2 Maintenance Training

Maintenance Technician Training Program:

Module	Duration	Content
Electrical Vehicle Systems	16 hours	High-voltage safety, battery management, charging systems, 12V/24V systems
Sensor Systems	16 hours	LiDAR maintenance and cleaning, camera systems, radar, GNSS, sensor replacement procedures
Calibration	8 hours	Intrinsic and extrinsic calibration procedures, calibration validation, tools and targets
Compute Systems	8 hours	Compute hardware, storage, networking, software update procedures, log extraction
Mechanical Systems	8 hours	Brakes, steering, suspension, tires, body, cargo interface
Diagnostic Tools	8 hours	Vehicle diagnostic interface, health monitoring dashboard, common fault codes, troubleshooting trees
Safety Procedures	4 hours	Lockout/tagout, high-voltage isolation, safe lifting, PPE requirements

Total Duration: 68 hours (approximately 8.5 days)

9.3 Incident Response Training

For all AV operations personnel (operators, maintenance, managers):

4-hour initial incident response training
Tabletop exercise simulating Level 1 and Level 2 incidents
Practical exercise with controlled scenario annually
Joint exercise with airport fire service annually
Debrief and lessons learned after every real incident

9.4 Airport Staff Awareness

For all airport personnel who may encounter AVs (ground handlers, airline staff, security, fire service):

Awareness Briefing (30-60 minutes):

What the AV looks like and where it operates
How the AV behaves (what to expect: speed, sounds, lights, yielding behavior)
What the AV can and cannot detect
How to behave around the AV (do not obstruct, maintain distance, use designated crossings)
How to identify if the AV is in autonomous or manual mode (indicator lights)
What to do if the AV behaves unexpectedly (do not approach, contact Airport Ops)
Emergency stop: location and use of external E-Stop button (if equipped)
Contact information for AV operations team

Delivery Method:

Include in existing airside safety induction for new personnel
Distribute as safety briefing document to all affected organizations
Post awareness signage at AV route crossing points
Brief via toolbox talks before AV operations begin in a new area
Annual refresher included in airside safety recertification

10. Maintenance and Support

10.1 Preventive Maintenance Schedule

Task	Frequency	Duration	Performed By
Daily CIL (Clean, Inspect, Lubricate)	Every shift start	15-30 min	Operator
Sensor lens/window cleaning	Daily (more in dusty/wet conditions)	10 min	Operator
Visual inspection of body, tires, lights	Daily	10 min	Operator
Check for FOD damage, fluid leaks	Daily	5 min	Operator
E-Stop function test	Daily	2 min	Operator
Weekly Checks	Weekly	1-2 hours	Technician
Tire pressure and condition check	Weekly	15 min	Technician
Brake system inspection	Weekly	20 min	Technician
Battery health check (state of health, cell balance)	Weekly	15 min	Technician
Communication system test (all links)	Weekly	15 min	Technician
Data download and log review	Weekly	30 min	Technician
Monthly Checks	Monthly	4-6 hours	Technician
Sensor calibration verification	Monthly	2 hours	Technician
Full extrinsic calibration check	Monthly	1 hour	Technician
Steering alignment check	Monthly	30 min	Technician
Suspension inspection	Monthly	30 min	Technician
Wiring and connector inspection	Monthly	30 min	Technician
Software diagnostic review	Monthly	30 min	Technician
Cleaning of all sensor apertures with approved materials	Monthly (deep clean)	30 min	Technician
Quarterly Service	Quarterly	Full day	Technician
Full brake service (pads, fluid, lines)	Quarterly	2 hours	Technician
Full calibration (intrinsic + extrinsic)	Quarterly	3 hours	Technician
Battery deep diagnostic	Quarterly	1 hour	Technician
Charging system inspection	Quarterly	1 hour	Technician
HD map accuracy validation drive	Quarterly	2 hours	Technician + Operator
Semi-Annual Service	Every 6 months	1-2 days	Technician
Full vehicle service per manufacturer schedule	Semi-annual	Full day	Technician
Sensor hardware health assessment	Semi-annual	4 hours	Technician
Compute hardware inspection (fans, thermal paste, storage health)	Semi-annual	2 hours	Technician
Safety system comprehensive test	Semi-annual	4 hours	Technician
Annual Service	Annually	2-3 days	Technician + Specialist
Full vehicle inspection and certification	Annual	Full day	Technician
Independent safety audit	Annual	1-2 days	External auditor
Comprehensive sensor replacement assessment	Annual	4 hours	Specialist
Insurance renewal inspection	Annual	Half day	Technician + Insurer

10.2 Sensor Cleaning

Cleaning Materials:

Denatured alcohol or sensor-specific cleaning solution (manufacturer approved)
Lint-free microfiber cloths
Compressed air (oil-free, filtered)
Soft-bristle brush for loose debris removal

Do NOT Use:

Paper towels, tissue, or abrasive cloths
Household glass cleaners (may damage optical coatings)
High-pressure water jets on sensor apertures
Solvents not approved by sensor manufacturer

Automated Cleaning Systems:

Consider integrating automated lens cleaning systems that use fluid-based nozzles or mechanical wipers
Automated systems trigger when 15-30% dirt coverage is detected
Image processing software should differentiate between rain, dust, and debris for targeted cleaning

Cleaning Triggers:

Scheduled: at start of each shift
Condition-based: when perception system reports degraded detection confidence (<95% target)
Weather-driven: after rain, dust storm, snow, or bird strike
Incident-driven: after any off-normal event

10.3 Calibration Checks

Field Calibration Validation (Monthly):

Drive the calibration validation route (defined section of operational route with known reference points)
Verify LiDAR point cloud alignment between multiple sensors (<2 cm disagreement)
Verify LiDAR-to-camera projection accuracy (❤️ pixels at 50 m)
Verify localization accuracy against ground truth markers (<10 cm)
Check object detection confidence at known distances (>95% at operational range)
Document results in calibration log

Full Recalibration Triggers:

Any physical impact to the vehicle
Sensor replacement
Monthly validation shows drift beyond threshold
After any major maintenance involving sensor mounting

10.4 Software Updates

OTA Update Process:

Software team develops and tests update in simulation and staging environments
Update package is cryptographically signed and validated (per UN R155/R156)
Update is deployed to 1 vehicle first (canary deployment)
Canary vehicle runs in shadow mode for 24-48 hours with new software
If no issues: deploy to remaining fleet in rolling update (one vehicle at a time)
Each vehicle runs in shadow mode for 4 hours after update before returning to autonomous operations
Rollback capability: ability to revert to previous software version within 30 minutes

Update Categories:

Critical safety patch: Deploy within 24 hours, may require operational pause
Bug fix: Deploy within 1 week
Feature enhancement: Deploy in next scheduled maintenance window
HD map update: Deploy within 48 hours of map change, shadow mode validation required

Geofenced Rollout:

Updates are first tested on a small group of vehicles and gradually rolled out
Monitoring period between batches to verify stability
Full fleet update within 2 weeks of initial canary deployment

11. Operational Procedures

11.1 Daily Pre-Flight Checks

Pre-Shift Vehicle Inspection Checklist (15-30 minutes per vehicle):

Exterior Inspection:

[ ] Walk around vehicle: check for damage, leaks, FOD attached to vehicle
[ ] Tires: visual condition check, no cuts or bulges
[ ] Lights: headlights, taillights, turn signals, hazard lights, beacon -- all functional
[ ] Sensors: all sensor lenses/windows clean and undamaged
[ ] Sensor mounts: secure, no visible misalignment
[ ] Body panels: secure, no loose parts
[ ] Cargo area / dolly interface: secure, functional
[ ] Charging connector: undamaged, clean
[ ] E-Stop buttons: accessible, not damaged, covers in place

Interior / System Checks:

[ ] Battery State of Charge: above minimum threshold for planned missions (typically >40%)
[ ] Compute system: powered on, all processes running, no error alerts
[ ] Sensor health dashboard: all sensors reporting, no degradation warnings
[ ] Communication: cellular signal confirmed, ROC connection verified
[ ] GNSS: position fix acquired, RTK corrections active, position accuracy <5 cm
[ ] Localization: converged on HD map, localization confidence >95%
[ ] E-Stop test: activate and deactivate hardware E-Stop, verify system response
[ ] Geofence: verify current geofence configuration loaded
[ ] Software version: matches approved operational version
[ ] NOTAM check: review current NOTAMs for route impact (automated or manual)

Confirmation:

[ ] Pre-shift inspection recorded in fleet management system
[ ] Any defects reported and categorized (ground vehicle / return to service)
[ ] Vehicle status set to "Ready for Operations" or "Not Available" with reason

11.2 Start-Up / Shutdown Procedures

Start-Up Sequence:

Complete pre-shift inspection checklist
Log into fleet management system with operator credentials
Verify ROC operator is on duty and monitoring
Confirm with Airport Operations that AV operations are clear to commence
Select first mission from dispatch queue
Verify route is clear (check for NOTAMs, construction, known obstructions)
Confirm weather conditions within operational limits
Transition vehicle to autonomous mode
Monitor first 2 minutes of operation closely for any anomalies
Confirm successful mission start to ROC

Shutdown Sequence:

Complete current mission or navigate to designated safe parking area
Transition vehicle to manual mode
Park vehicle in designated position
Apply parking brake
If charging needed: connect to charger, verify charging initiated
Run post-trip inspection:
- [ ] New damage check (compare to pre-trip condition)
- [ ] Fluid leak check
- [ ] Sensor condition check
- [ ] Log any anomalies observed during operations
Upload remaining data logs
Set vehicle status to "Parked" or "Charging" in fleet management system
Complete shift report
Hand over to incoming operator (if shift change) with verbal briefing on:
- Vehicle condition
- Any issues encountered
- Remaining mission queue
- Weather forecast for next shift

11.3 Emergency Procedures

AV Emergency Stop:

Activate nearest E-Stop (hardware button on vehicle, remote from ROC, or software)
Vehicle stops and engages parking brake
Hazard lights activate automatically
Notify Airport Ops immediately via radio
Assess the situation from a safe distance
If safe to approach: deploy warning cones around vehicle
Follow incident response procedure based on severity level

Communication Loss:

Vehicle automatically executes safe-stop after 5-second timeout
ROC operator alerts field support
Field support travels to vehicle location
Field support assesses communication status
If comms restored: resume operations from ROC after system health check
If comms not restored: manual drive to maintenance area

Vehicle Fire:

Activate E-Stop (if not already)
Evacuate all personnel from 50 m radius
Notify Airport Fire Service immediately
Do NOT attempt to fight a battery fire with water (use Class D extinguisher or dry chemical)
Airport Fire Service responds per airport emergency plan
AV operations team provides vehicle technical briefing to fire crew (battery location, high-voltage isolation procedure)

Aircraft Emergency on Airport:

ROC operator receives airport emergency notification
All AVs immediately execute safe-stop and move to shoulder/edge of route
AVs must not impede emergency vehicle access routes
Operations remain suspended until airport declares "all clear"
Post-emergency: verify all AVs safe and operational before resuming

Adverse Weather Procedures:

Condition	Threshold	Action
Rain (light)	<10 mm/h	Normal operations, increased sensor cleaning
Rain (moderate)	10-25 mm/h	Reduce speed by 30%, increase following distance
Rain (heavy)	25-50 mm/h	Reduce speed by 50%, restrict to essential routes
Rain (extreme)	>50 mm/h	Suspend operations, vehicles safe-stop
Fog	Visibility 500-1000 m	Reduce speed by 30%
Fog (dense)	Visibility <500 m	Suspend operations
Wind	30-50 km/h	Reduce speed by 20%, monitor stability
Wind (strong)	>50 km/h	Suspend operations
Snow / Ice	Any accumulation on route	Suspend operations (unless specifically validated)
Thunderstorm	Within 10 km	Suspend operations, vehicles safe-stop in protected area
Dust storm	Visibility <500 m	Suspend operations

Weather Monitoring:

Integrate with airport METAR/AWOS feeds for automated weather monitoring
Define automatic weather alerts in fleet management system
ROC operator has authority to suspend operations for weather at any time
Operations resume only after weather conditions return to limits for 30 minutes

12. KPIs and Reporting

12.1 What to Measure

Safety KPIs:

KPI	Definition	Target (Pilot)	Target (Operations)
Safety Disengagement Rate	Safety disengagements per 1,000 autonomous km	<2.0	<0.2
Non-Safety Disengagement Rate	Non-safety disengagements per 1,000 autonomous km	<10.0	<1.0
Collision Rate	Collisions per 100,000 autonomous km	0	0
Near-Miss Rate	Near-miss events per 1,000 autonomous km	<5.0	<1.0
E-Stop Activations	E-Stop events per 1,000 autonomous km	<1.0	<0.1
Geofence Violations	Soft boundary alerts per 1,000 autonomous km	<0.5	<0.1
Incident Count	Level 1-2 incidents per month	0	0
Mean Distance Between Disengagements	Autonomous km per disengagement	>500 km	>5,000 km

Operational KPIs:

KPI	Definition	Target
System Availability	% of scheduled operating time vehicle is available	>90% (pilot), >95% (ops)
Mission Completion Rate	% of assigned missions completed successfully	>95% (pilot), >99% (ops)
On-Time Performance	% of missions completed within schedule window	>85% (pilot), >95% (ops)
Average Speed	Average moving speed as % of allowed speed	>70%
Vehicle Utilization	% of available hours vehicle is on mission	>40% (pilot), >70% (ops)
Communication Availability	% of time at least one comms link is active	>99.0% (pilot), >99.9% (ops)
Charging Efficiency	% of charging sessions completing without issue	>95%
Localization Accuracy	95th percentile position error (cm)	<15 cm

Maintenance KPIs:

KPI	Definition	Target
Preventive Maintenance Compliance	% of scheduled PM tasks completed on time	>95%
Mean Time Between Failures (MTBF)	Average hours between unplanned maintenance	>200 hours (pilot), >500 hours (ops)
Mean Time to Repair (MTTR)	Average hours to return vehicle to service	<4 hours
Sensor Replacement Rate	Sensor replacements per vehicle per quarter	<2
Software Update Success Rate	% of OTA updates applied successfully	>99%

12.2 How to Report to Airport Authority

Monthly Operational Report (Required):

Executive Summary (1 page)
- Key metrics dashboard
- Notable achievements and milestones
- Issues and remediation status
Safety Performance (2-3 pages)
- All KPIs from safety table above, with trend graphs
- All incidents and near-misses with root cause analysis
- Risk register updates
- Safety improvement actions taken
Operational Performance (2-3 pages)
- Autonomous km and hours driven
- Mission volume and completion rates
- Route performance comparison
- Weather-related operational impacts
Maintenance and Reliability (1-2 pages)
- Vehicle availability and downtime causes
- Maintenance activities completed
- Software updates deployed
- Hardware issues and replacements
Upcoming Plans (1 page)
- Next month's operational plan
- Route or fleet changes planned
- Software updates scheduled
- Training activities planned

Quarterly Safety Review (Formal Presentation):

Present to Airport Safety Committee
Comprehensive safety trend analysis
Risk register review (all items)
Safety case update
External incident benchmarking
Safety improvement roadmap

12.3 Continuous Improvement Loop

Data-Driven Improvement Cycle:

Collect Data --> Analyze Performance --> Identify Gaps -->
Root Cause Analysis --> Implement Changes --> Validate Impact -->
Update Procedures --> Collect Data (repeat)

Key Improvement Processes:

Daily Data Review: Operations team reviews all disengagements, near-misses, and anomalies from previous 24 hours. Categorize by root cause. Feed into weekly analysis.
Weekly Performance Review: Operations manager reviews KPI trends, identifies degradation, prioritizes improvements. Feed critical issues to engineering team.
Monthly Safety Review: Safety manager reviews all safety KPIs, incident trends, risk register. Prepare monthly report for airport authority.
Quarterly Strategic Review: Leadership reviews overall program performance, business case tracking, expansion readiness, stakeholder satisfaction.
Continuous Scenario Learning: Every disengagement and near-miss becomes a new test scenario. Replay in simulation. Validate software fix. Add to regression test suite. Deploy improvement. This creates a virtuous cycle where operational experience directly improves system capability.

13. Multi-Airport Expansion

13.1 Adapting to New Airports

New Airport Assessment Checklist:

Assessment Area	Key Questions
Regulatory	Different country/jurisdiction? What aviation authority? What autonomous vehicle regulations?
Infrastructure	GNSS coverage quality? Cellular coverage? Power availability? Weather conditions?
Layout	Apron configuration? Gate types? Vehicle traffic patterns? Surface conditions?
Operations	Different ground handlers? Different airlines? Different flight mix? Different peak patterns?
Climate	Temperature extremes? Precipitation types and frequency? Wind patterns? Visibility patterns?
Stakeholders	Different airport authority expectations? Different ATC procedures? Different security requirements?

13.2 What's Reusable

Component	Reusability	Notes
Vehicle hardware	100% reusable	Same vehicle operates at any airport (adjust for local regulations)
Perception software	95% reusable	Core algorithms transfer; may need retraining for local GSE types, aircraft mix
Planning software	90% reusable	Core planner transfers; speed profiles and intersection logic may need adjustment
Control software	95% reusable	Vehicle dynamics same; surface conditions may differ
Fleet management system	95% reusable	Same platform; configure for local routes and schedules
ROC systems	90% reusable	Same interface; configure for local operations
Safety case framework	80% reusable	Same methodology; hazard analysis must be site-specific
Training curriculum	85% reusable	Same core content; airport-specific module must be updated
Maintenance procedures	95% reusable	Same vehicle, same procedures
Operational procedures	70% reusable	Core procedures transfer; site-specific procedures needed
KPI framework	95% reusable	Same metrics; targets may differ based on site complexity
Insurance framework	60% reusable	Policy structure reusable; coverage and terms site-specific

13.3 What Needs to Be Redone

Activity	Effort (% of first site)	Duration
Site survey	100%	1-2 weeks
HD mapping	100%	1-2 weeks
Waypoint creation and geofencing	100%	1-2 weeks
Stakeholder engagement	80%	2-3 months
Airport authority approval	80%	2-4 months
Risk assessment (site-specific)	70%	2-4 weeks
SAT testing	60%	1-2 weeks (streamlined from first site)
Pilot operations	50%	3-6 months (can be accelerated with proven track record)
Training (airport-specific module)	30%	1 week
Communication setup	80%	1-2 weeks
Charging infrastructure	100%	4-8 weeks

Expected Timeline for Second Airport: 6-9 months from site survey to routine operations (compared to 12-18 months for first airport)

Expected Timeline for Third+ Airport: 4-6 months (further streamlined with reusable tooling and experienced team)

13.4 Multi-Airport Fleet Management

Centralized ROC capable of monitoring vehicles across multiple airports
Standardized software deployment pipeline across all sites
Shared incident database and lessons-learned repository
Cross-airport performance benchmarking
Centralized spare parts inventory with logistics to each site
Traveling deployment team for new site setup
Site-specific operations teams for day-to-day management

14. Insurance and Liability

14.1 Who's Liable

Liability Framework for Airport AV Operations:

Scenario	Primary Liability	Contributing Liability
AV collides with aircraft	AV operator/deployer	AV manufacturer (if system defect)
AV collides with person	AV operator/deployer	AV manufacturer (if perception failure)
AV collides with vehicle	AV operator/deployer	Other vehicle operator (if at fault)
AV collides with infrastructure	AV operator/deployer	Airport (if infrastructure defect)
Software defect causes incident	AV manufacturer	AV operator (if known defect, failed to update)
Sensor failure causes incident	Sensor manufacturer	AV integrator (if integration defect)
Cyber attack causes incident	AV operator (duty of care)	Attacker (criminal liability)
Cargo damage during transport	AV operator	Ground handler (if loading error)
Operator error (wrong mode, wrong route)	AV operator	Training provider (if inadequate training)

Key Liability Principles:

The entity deploying the AV (typically the airport operator or the AV service provider) holds primary operational liability
Product liability may shift to the AV manufacturer for defects in design, manufacturing, or failure to warn
Joint liability may apply when multiple parties contribute to an incident
Strict liability applies in some jurisdictions: the AV operator is liable regardless of fault
Contractual allocation of liability between airport and AV provider should be clearly defined in the operating agreement

14.2 Insurance Requirements

Required Insurance Policies:

Policy Type	Purpose	Recommended Coverage
Commercial General Liability (CGL)	Third-party bodily injury and property damage	$10-25 million per occurrence
Automobile Liability	Vehicle-specific liability for AV operations	$5-10 million per occurrence
Product Liability	Defects in AV technology	$10-25 million aggregate (held by manufacturer)
Cyber Liability	Data breach, cyber attack, system compromise	$5-10 million per occurrence
Airport Operator's Liability	Airport-specific coverage (held by airport)	Per airport's existing policy
Professional Liability (E&O)	Errors in AV system design, deployment, operations	$5-10 million aggregate
Workers' Compensation	Injury to AV operators and technicians	Per local statutory requirements
Property / Equipment	AV vehicle damage, sensor damage, compute equipment	Replacement value per vehicle
Business Interruption	Loss of revenue due to operational suspension	6-12 months operating costs

Note on Coverage Amounts:

Coverage amounts vary significantly based on airport size, traffic volume, and jurisdiction
California requires $5 million minimum for AV testing; Florida requires $1 million minimum
Airport operators typically require tenants to carry $10-25 million CGL as standard
Commercial AV operators often carry $5 million or more per vehicle
Consult with an aviation insurance broker specializing in airport operations for site-specific guidance

14.3 Claims Handling

Claims Process:

Immediate: Activate incident response procedure (see Section 6.3)
Within 24 hours: Notify insurance provider with preliminary incident report
Within 48 hours: Preserve all evidence (vehicle data, sensor logs, CCTV footage, witness statements)
Within 7 days: Submit formal claim with detailed incident report and supporting documentation
Investigation: Cooperate fully with insurer's investigation team; provide data access
Resolution: Work with insurer and legal counsel to resolve claims

Data Preservation for Claims:

All vehicle sensor data (LiDAR, camera, radar) for 1 hour before and after incident
All vehicle telemetry data for 24 hours before and after incident
All communication logs between vehicle, ROC, and field personnel
All system health data and error logs
Airport CCTV footage (request from airport security within 24 hours)
Weather data at time of incident
NOTAM data at time of incident
HD map version and geofence configuration at time of incident

Retain all incident data for minimum 7 years (or longer per local statute of limitations for personal injury claims).

15. Master Deployment Checklist

Phase 0: Preparation (Months 1-4)

Business Case and Strategy:

[ ] Build business case with ROI model
[ ] Identify executive sponsor
[ ] Secure initial funding
[ ] Select AV technology partner/vendor
[ ] Define target operational design domain (routes, areas, times)

Regulatory Engagement:

[ ] Identify applicable aviation authority (FAA, CAA, EASA, etc.)
[ ] Review FAA CertAlert 24-02 and Bulletin 25-02 (US) or equivalent guidance
[ ] Contact FAA Regional Airport Certification Safety Inspector (Part 139) or Airports District Office
[ ] Understand approval pathway and timeline
[ ] Engage legal counsel on regulatory compliance

Site Survey:

[ ] Complete physical environment assessment
[ ] Complete operational environment assessment
[ ] Complete infrastructure assessment
[ ] Identify communication coverage gaps
[ ] Document GPS shadow zones
[ ] Produce site survey report

Stakeholder Engagement:

[ ] Conduct internal alignment meeting
[ ] Brief airport authority leadership
[ ] Engage FAA/CAA informally
[ ] Identify all affected external stakeholders
[ ] Create stakeholder communication plan
[ ] Begin airline and ground handler engagement

Phase 1: Mapping and Vehicle Preparation (Months 4-7)

Mapping:

[ ] Establish ground control point reference network
[ ] Conduct mapping data collection drives
[ ] Process point cloud data
[ ] Extract and annotate features
[ ] Build HD map (Road Model, Lane Model, Localization Model)
[ ] Validate HD map with test drives
[ ] Define waypoints for all planned routes
[ ] Implement geofence boundaries (hard, soft, speed, dynamic)
[ ] Set up NOTAM integration

Vehicle Preparation:

[ ] Complete sensor mounting and wiring
[ ] Perform intrinsic calibration of all sensors
[ ] Perform extrinsic calibration (sensor-to-sensor, sensor-to-vehicle)
[ ] Install compute hardware with safety architecture
[ ] Set up communication stack (cellular, Wi-Fi, V2X)
[ ] Install and test E-Stop systems (all types)
[ ] Conduct Factory Acceptance Test (FAT)
[ ] Ship vehicle to airport site

Infrastructure:

[ ] Install charging infrastructure
[ ] Set up edge compute / communication base stations (if needed)
[ ] Set up Remote Operations Center (ROC)
[ ] Verify communication coverage across operational area
[ ] Install any required physical infrastructure (signage, markings, barriers)

Phase 2: Testing (Months 7-9)

Site Acceptance Test:

[ ] Complete SAT Phase 1: Static tests
[ ] Complete SAT Phase 2: Low-speed controlled driving
[ ] Complete SAT Phase 3: Operational speed testing
[ ] Complete SAT Phase 4: Stress testing
[ ] Document all SAT results

Integration Testing:

[ ] Test fleet management system integration
[ ] Test A-CDM / flight data integration
[ ] Test NOTAM integration
[ ] Test charging system integration
[ ] Test ROC monitoring and control
[ ] Test data pipeline end-to-end

Safety Preparation:

[ ] Complete Hazard Identification (HAZID) workshop
[ ] Complete Safety Risk Assessment (SRA)
[ ] Create risk register with all identified hazards
[ ] Develop incident response plan
[ ] Develop emergency procedures
[ ] Conduct tabletop emergency exercise
[ ] Compile safety case document

Approvals:

[ ] Submit safety case to airport authority
[ ] Obtain written approval for pilot operations
[ ] Confirm FAA/CAA notification complete
[ ] Confirm insurance coverage active
[ ] Confirm airline and ground handler acknowledgment

Phase 3: Training (Months 8-9)

[ ] Complete operator training (96 hours per operator)
[ ] Complete maintenance technician training (68 hours per technician)
[ ] Complete incident response training (all personnel)
[ ] Conduct airport staff awareness briefings
[ ] Brief airport fire service on AV emergency procedures
[ ] Certify all operators (written + practical exam)
[ ] Brief ATC on AV operations and coordination protocol

Phase 4: Pilot Operations (Months 9-18)

[ ] Conduct Operational Readiness Review (ORR)
[ ] All ORR criteria passed: formal Go decision
[ ] Begin Phase A: Shadow mode (4 weeks)
[ ] Review shadow mode data, confirm readiness
[ ] Begin Phase B: Supervised autonomy (12 weeks)
[ ] Achieve disengagement rate target
[ ] Airport authority review at mid-pilot point
[ ] Begin Phase C: Remote supervision (10 weeks)
[ ] Demonstrate reliability for remote supervision reduction
[ ] Begin Phase D: Routine operations (26 weeks)
[ ] Submit monthly operational reports to airport authority
[ ] Conduct quarterly safety reviews

Phase 5: Scaling (Months 18+)

Fleet Expansion:

[ ] Achieve target KPIs for 30 consecutive days
[ ] Obtain airport authority approval for expansion
[ ] Scale ROC staffing for expanded fleet
[ ] Scale maintenance capacity
[ ] Scale charging infrastructure
[ ] Update insurance coverage
[ ] Add vehicles per expansion schedule

Route Expansion:

[ ] Map and validate new routes
[ ] Conduct SRA for new routes
[ ] Complete abbreviated SAT on new routes
[ ] Shadow mode on new routes
[ ] Transition to supervised, then routine operations

Supervision Reduction:

[ ] Achieve supervision reduction metrics
[ ] Obtain approvals (Safety Manager, Airport Ops, Airport Safety, FAA/CAA)
[ ] Implement new supervision model
[ ] Monitor closely for 30 days after change
[ ] Document results and update safety case

Continuous Improvement:

[ ] Daily data review and anomaly categorization
[ ] Weekly performance review
[ ] Monthly safety and operational reports
[ ] Quarterly strategic reviews
[ ] Annual safety audit
[ ] Annual insurance review
[ ] Ongoing scenario learning from operational data

Phase 6: Multi-Airport Expansion (When Ready)

[ ] Complete new airport assessment checklist
[ ] Identify site-specific requirements and delta from first site
[ ] Engage new airport authority and regulator
[ ] Conduct site survey and mapping
[ ] Deploy vehicles to new site
[ ] Execute streamlined testing and pilot program
[ ] Establish local operations team
[ ] Integrate into centralized fleet management
[ ] Begin operations (target: 6-9 months for second site, 4-6 months for subsequent)

Appendix A: Key Regulatory References

Reference	Title	Relevance
FAA CertAlert 24-02	Autonomous Ground Vehicle Systems (AGVS) Technology on Airports	Primary US guidance on AGVS at airports
FAA Bulletin 25-02	AGVS Testing in Closed Movement Areas	US guidance on testing in movement areas
AC 150/5210-20A	Ground Vehicle Operations on Airports	General airport ground vehicle rules
AC 150/5220-26	Airport Ground Vehicle Automatic Dependent Surveillance	VMAT transponder requirements
14 CFR Part 139	Certification of Airports	Airport certification standards
ICAO Annex 14	Aerodromes	International aerodrome standards
IATA AHM 908	Autonomous Vehicle Standards	Ground handling autonomous vehicle protocols
SAE J3016	Taxonomy of Automated Driving	Automation level definitions
ISO 26262	Functional Safety for Road Vehicles	Safety engineering standard
ISO 21448 (SOTIF)	Safety of the Intended Functionality	Addresses limitations and foreseeable misuse
UL 4600	Safety for Autonomous Products	Safety case framework for autonomous systems
UN R155/R156	Cybersecurity and Software Updates	OTA update security requirements
ACRP Research Report 219	Advanced Ground Vehicle Technologies for Airside Operations	Comprehensive ACRP research on airside AV

Appendix B: Key Industry Deployments (Reference Cases)

Airport	Technology	Operator	Status	Key Learning
Cincinnati/Northern Kentucky (CVG)	autonomous baggage/cargo tug (reference airside AV stack)	CVG + airlines	Operational testing	2,163 km over 3 weeks; 81.3% zero-disengagement mission rate; 13.2 km per disengagement
Singapore Changi	autonomous baggage/cargo tug (reference airside AV stack)	Changi Airport Group	Phase 2 trials (at aircraft stand)	Rain-sensing algorithm validated to 50 mm/h; multi-phase testing approach
Teesside International	autonomous baggage/cargo tug + passenger shuttle	reference airside AV stack	Testing from Jan 2026, airside from 2027	World's first combined cargo + passenger autonomous airport vehicles
Amsterdam Schiphol	Autonomous baggage tug	Multiple trials	Advanced trials	Large hub complexity; mixed traffic management
Dubai (DXB)	TractEasy autonomous tug	dnata + EasyMile	Level 3 operational, Level 4 from 2026	Created new UAE regulatory framework for airport AV; GCAA collaboration
East Midlands (UK)	autonomous baggage/cargo tug (reference airside AV stack)	reference airside AV stack	Ground handling license issued	Proving ground strategy before multi-site expansion
Phoenix Sky Harbor	Waymo robotaxi (landside)	Waymo + PHX	Operational	Landside rideshare to terminal; different use case but demonstrates airport AV acceptance

Appendix C: Glossary

Term	Definition
A-CDM	Airport Collaborative Decision Making
ADS-B	Automatic Dependent Surveillance-Broadcast
AGVS	Autonomous Ground Vehicle Systems
AHM	Airport Handling Manual (IATA)
ASIL	Automotive Safety Integrity Level
ATC	Air Traffic Control
AWOS	Automated Weather Observing System
CIL	Clean, Inspect, Lubricate
CONOPS	Concept of Operations
C-V2X	Cellular Vehicle-to-Everything
DSRC	Dedicated Short-Range Communications
E-Stop	Emergency Stop
FAT	Factory Acceptance Test
FIDS	Flight Information Display System
FOD	Foreign Object Debris
GNSS	Global Navigation Satellite System
GSE	Ground Support Equipment
HAZID	Hazard Identification
HD Map	High-Definition Map
ILS	Instrument Landing System
IMU	Inertial Measurement Unit
INS	Inertial Navigation System
LiDAR	Light Detection and Ranging
METAR	Meteorological Terminal Air Report
MTBF	Mean Time Between Failures
MTTR	Mean Time to Repair
NOTAM	Notice to Air Missions
ODD	Operational Design Domain
ORR	Operational Readiness Review
OTA	Over-the-Air (software update)
ROC	Remote Operations Center
RTK	Real-Time Kinematic (GNSS corrections)
SAE	Society of Automotive Engineers
SAT	Site Acceptance Test
SLAM	Simultaneous Localization and Mapping
SMS	Safety Management System
SOTIF	Safety of the Intended Functionality
SRA	Safety Risk Assessment
V2X	Vehicle-to-Everything
VMAT	Vehicle Movement Area Transmitter
vSOC	Vehicle Security Operations Center

This playbook is based on research conducted in March 2026, incorporating guidance from the FAA, ICAO, IATA, and lessons learned from deployments at CVG, Singapore Changi, Teesside, Dubai, Schiphol, and East Midlands airports. Regulatory frameworks for airport autonomous vehicles are evolving rapidly. Always verify current requirements with your aviation authority before proceeding.

SLAM Methods

Methods

Deployment Playbook: Autonomous Vehicles at Airports ​

From First Test to Full Operations ​

Table of Contents ​

1. Pre-Deployment Phase ​

1.1 Site Survey ​

1.2 Stakeholder Engagement ​

1.3 Airport Authority Meetings ​

1.4 Safety Assessment ​

1.5 Risk Register ​

2. Mapping and Route Planning ​

2.1 How to Map an Airport for Autonomous Operations ​

2.2 Waypoint Creation ​

2.3 Geofencing ​

2.4 NOTAM Integration ​

3. Vehicle Preparation ​

3.1 Sensor Mounting ​

3.2 Calibration ​

3.3 Compute Installation ​

3.4 Communication Setup ​

3.5 Safety Systems Validation ​

4. Testing Phases ​

4.1 Factory Acceptance Test (FAT) ​

4.2 Site Acceptance Test (SAT) ​

4.3 Integration Testing ​

4.4 Operational Readiness Review (ORR) ​

5. Pilot Operations ​

5.1 How to Run a Pilot (1-3 Vehicles) ​

5.2 Scope Limitations ​

5.3 Human Supervision Requirements ​