Industry Research

Appearance

Singapore Changi Airport: Autonomous Ground Support Equipment Program ​

The Most Advanced Airside Autonomy Deployment Globally ​

Singapore Changi Airport operates the most comprehensive autonomous ground support equipment (GSE) program at any international airport outside China. With multiple concurrent autonomous vehicle programs spanning baggage tractors, cargo dollies, and staff buses, Changi has established itself as the global reference site for airside autonomy -- advancing from early proof-of-concept trials in 2020 to fully driverless live operations by January 2026.

1. Full Program History and Timeline ​

2017: Genesis ​

  • October 2017: TLD (GSE manufacturer, part of Alvest Group) and EasyMile (autonomous driving software company) announce partnership to develop TractEasy, the first autonomous baggage tractor. The vision is to commercialize the EZTow autonomous tow tractor for airport use.

2020: First On-Ground Testing ​

  • October 2020: The TractEasy EZTow is tested at Changi Terminal 4, which had suspended flights since May 2020 due to COVID-19. This sterile environment allowed testing of route-following capabilities without interference from other vehicles.
  • Testing subsequently moved to Terminal 3 for "live" environment trials with moving obstacles and other airside traffic.

2021: Live Flight Trials Begin ​

  • June 2021: TractEasy is formally launched as a joint venture between Alvest Group (through TLD and Smart Airport Systems/SAS) and EasyMile.
  • August 2021: Changi Airport Group (CAG) partners with SATS to trial TractEasy for autonomous baggage towing on live flights, with CAAS support. This is the first autonomous vehicle initiative approved for airside operations at Changi.
  • CAG planned to add two additional autonomous baggage tractors to the trial by October 2021.

2022: reference airside AV stack Partnership Begins ​

  • October 2022: CAG signs initial agreement with UK-based reference airside AV vendor to trial their autonomous baggage/cargo tug autonomous baggage handling vehicle.
  • February 2023: Multi-year formal partnership agreement signed for continued reference airside AV stack development and testing.
  • the reference airside AV stack's Phase 2A trials begin, running for two years to test the autonomous baggage/cargo tug in wet weather, humidity, alignment with airport equipment, and interaction with traditional GSE.

2023: 5G Testbed and Expanded Trials ​

  • March 2023: CAAS and Singtel launch the 5G Aviation Testbed at Terminal 3 airside, enabling real-time teleoperation of autonomous vehicles via high-definition video streams.
  • May 2023: CAAS issues Advisory Circular AC-139-7-7, the first formal regulatory guidance for autonomous vehicles at airside in Singapore.
  • July 2023: Official public launch of the 5G Aviation Testbed, with partners including CAAS, Singtel, CAG, Singapore Airlines, and reference airside AV stack.
  • Three TractEasy EZTow units are operational, transporting baggage between a terminal and its satellite building with only remote supervision.
  • Trials with UISEE autonomous tractors also underway.

2024: Infrastructure Investment and New Programs ​

  • CAG, Nokia, and M1 Limited roll out a dedicated private 5G network for airside operations, supporting autonomous ground vehicles, baggage handling, and drone-based maintenance inspections.
  • May 2024: reference airside AV stack announces deployment of four new autonomous baggage/cargo tug vehicles for Phase 2B underwing operations trials at Changi.
  • July 2024: SATS, SIA Engineering Company, and CAG sign MoU for autonomous airside bus trials (two-year program, co-funded by CAAS).
  • October 2024: SATS announces S$250 million investment to modernize Changi ground and cargo operations, including autonomous technology integration.
  • Late 2024 / November: UISEE autonomous tractor trials begin in earnest -- safety driver onboard for initial phase.

2025: Intensive Testing ​

  • UISEE autonomous tractors undergo nearly a year of rigorous testing with a safety driver, completing more than 5,000 test trips and accumulating over 20,000 km of accident-free operation.
  • November 2025: Safety driver removed; transition to fully driverless operation with remote supervision only.
  • Singtel extends 5G coverage to public areas across all Changi terminals by end of 2025.

2026: Live Driverless Operations Launch ​

  • January 20, 2026: Official launch of fully driverless autonomous tractor operations, officiated by Ms. Sun Xueling, Senior Minister of State for Transport. Two UISEE autonomous tractors begin live operations on the T1-T4 baggage route.
  • Later 2026 (planned): Six additional autonomous tractors to be deployed on a new route between Terminal 2's baggage handling area and aircraft stands, under the CAG-SATS collaboration.

2027: Fleet Scale-Up Target ​

  • Autonomous tractor fleet to expand to 24 vehicles, covering baggage, cargo, and equipment towing across the airport.

Key Partners Across the Program ​

PartnerRole
Changi Airport Group (CAG)Airport operator, program lead, infrastructure provider
SATS LtdGround handler, operational integration, Hub Handler of the Future programme
CAASRegulator, co-funder (Aviation Development Fund), safety approval
UISEE (China)Autonomous tractor technology provider (current live fleet)
EasyMile / TractEasy (France)Earlier autonomous tractor trials (EZTow), ongoing operations
TLD / Smart Airport Systems (France)GSE hardware manufacturer, TractEasy joint venture
reference airside AV vendor (UK)autonomous baggage/cargo tug autonomous baggage handling vehicle
NokiaPrivate 5G network infrastructure
M1 LimitedPrivate 5G deployment partner
Singtel5G Aviation Testbed provider
SIA Engineering CompanyAutonomous bus trial partner

2. Current Deployment (as of January 2026) ​

Operational Fleet ​

  • 2 autonomous tractors in live driverless operations (UISEE platform)
  • 3 TractEasy EZTow units operational in separate trial/operations (EasyMile platform)
  • 4 reference airside AV stack autonomous baggage/cargo tug vehicles in Phase 2B underwing trials

Primary Route: Terminal 1 to Terminal 4 ​

  • Distance: 7 km connecting the airport's oldest terminal (T1, opened 1981) and newest terminal (T4, opened 2017)
  • Function: Transferring passenger baggage between the baggage handling areas of T1 and T4
  • Operating conditions: Day, night, and rain -- all weather conditions
  • Supervision: Fully driverless with remote monitoring from a dedicated control center
  • Cargo per trip: Up to 4 unit loading devices (ULDs), combined weight up to 10 tonnes

Terminal Integration ​

  • Terminal 1: Baggage handling area (origin/destination point)
  • Terminal 2: Next deployment phase -- route between T2 baggage handling area and aircraft stands (6 additional tractors planned for later 2026)
  • Terminal 3: 5G Aviation Testbed location; earlier TractEasy trial operations
  • Terminal 4: Baggage handling area (origin/destination point on current T1-T4 route); site of original 2020 controlled testing

Operational Hours ​

  • The EZTow platform is designed for 24/7 duty cycles
  • UISEE systems are designed for continuous all-weather operation
  • Exact current shift coverage at Changi has not been publicly disclosed, but the system supports round-the-clock operations

3. Nokia Private 5G Network ​

Network Architecture ​

Changi Airport has deployed two distinct 5G initiatives:

1. Singtel 5G Aviation Testbed (2023-2025)

  • Partners: CAAS, Singtel, CAG
  • Coverage: Terminal 3 airside
  • Duration: Two-year trial from March 2023
  • Purpose: Enabling teleoperation of autonomous vehicles and secure flight data transfer
  • Expansion: Nearly 4,000 Singtel corporate mobile lines at airside upgraded to 5G

2. Nokia / M1 Private 5G Network (2024-present)

  • Partners: CAG, Nokia, M1 Limited
  • Technology: Nokia Digital Automation Cloud (DAC) platform
  • Purpose: High-capacity private 5G for autonomous ground vehicles, baggage handling systems, drone-based maintenance inspections, digital twin operations, and predictive maintenance
  • Architecture: Edge-centric with cloud support, ensuring data proximity and low latency for demanding applications

Measured Performance (Changi 5G Testing) ​

From Changi Airport's 5G testing with StarHub (Airport Emergency Services application):

BandDownlink (avg)Uplink (avg)Latency
700 MHz114.95 Mbps33.09 Mbps17 ms
700 MHz + CoW120.22 Mbps43.50 Mbps15 ms
3.5 GHz248.81 Mbps7.38 Mbps10 ms
3.5 GHz + CoW281.67 Mbps33.79 Mbps~10 ms

Video performance at range:

  • 4K video: 500 m with no pixelation
  • 2K video: 3 km
  • 1K video: 4.5 km

Why 5G Over WiFi ​

Private 5G was chosen over WiFi for airside autonomous vehicle operations for several critical reasons:

  1. Seamless mobility: WiFi suffers from latency spikes during handover between access points. 5G natively supports seamless mobility at speeds up to several hundred km/h, maintaining reliable connectivity for moving vehicles without interruption.

  2. Deterministic quality of service: WiFi uses Listen Before Talk (LBT) protocol -- if a channel is busy, devices wait, creating unpredictable delays. 5G provides deterministic QoS even in high-traffic areas with many concurrent users.

  3. Infrastructure reduction: 5G eliminates the need for extensive fiber optic cable runs to hundreds of WiFi access points across the apron, reducing deployment complexity and maintenance burden.

  4. Coverage of remote areas: 5G provides coverage to remote aircraft stands that were previously unsupported by wired infrastructure, critical for Singapore Airlines' flight data transfer application.

  5. Reliability in harsh environments: Airport airside presents challenges including electromagnetic interference from aircraft systems, large metal structures, and outdoor exposure. Private 5G provides carrier-grade reliability in these conditions.

  6. Video stream quality: Real-time HD video supervision of autonomous vehicles requires sustained high-bandwidth uplink with consistent low latency -- a combination WiFi cannot reliably deliver in outdoor environments.

4. Technical Details ​

UISEE Autonomous Tractor (Current Live Fleet) ​

Autonomous Driving Platform: UISEE U-Drive universal autonomous driving platform (SAE Level 4)

Three-Module Architecture:

  1. Intelligent driving controller (vehicle brain): Multi-sensor fusion positioning, perception, and control planning
  2. AI algorithm system: Real-time environmental perception, path planning, obstacle avoidance
  3. Cloud-based intelligent brain: Remote monitoring, vehicle dispatch, fleet management, big data analytics

Sensor Suite (varies by model):

  • 4-5 LiDAR sensors (Hesai XT series)
    • Ranging accuracy: 5 mm (1 sigma)
    • Max detection distance: 120 m
    • Zero blind range (detects objects directly touching sensor)
    • Service life: >30,000 hours
  • 6-7 high-definition cameras
  • RTK high-precision GNSS positioning
  • Inertial navigation system (IMU)
  • Centimeter-level positioning precision via multi-sensor fusion

Capabilities:

  • Autonomous driving, lane changing, obstacle avoidance, emergency braking
  • Precise docking at loading/unloading points
  • Geofencing and anti-tampering security measures
  • All-weather operation (rain, night, tropical humidity)

Certifications:

  • ISO 21434 Road Vehicle Cybersecurity Process Certification
  • ISO 27001 Information Security Management Certification
  • Singapore Technical Reference TR68 compliance

Connectivity: 4G/5G cellular, Road Side Units (RSU), WiFi/Ethernet to monitoring center

TractEasy EZTow (Earlier/Parallel Fleet) ​

Autonomous Driving: EasyMile driverless software stack (SAE Level 4)

Dimensions: 3,200 x 1,940 x 2,050 mm (L x W x H)

Performance:

  • Towing capacity: 14 tonnes (28,000 lbs) / drawbar pull: 2,000 daN (4,500 lbf)
  • Maximum speed: 15 km/h (9.3 mph) autonomous; typical operational speed ~10 km/h
  • Turning radius: 4,750 mm
  • Trailer capacity: Up to 4 ULDs; can pull 3 trailers with combined length >16 m

Power: Electric drive, lead-acid or lithium-ion battery options; designed for 24/7 duty cycle

Sensor Suite:

  • Multiple LiDAR sensors
  • Stereo cameras
  • Radar sensors
  • IMU (inertial measurement unit)
  • GPS / RTK positioning
  • Wheel encoders
  • 3G/4G modem
  • V2X on-board units
  • 360-degree surround perception

Navigation Approach:

  • GPS satellite navigation combined with LiDAR-based 3D mapping
  • EasyMile algorithm fuses all sensor data with GPS for centimeter-precision localization
  • Pre-mapped routes with real-time obstacle detection and avoidance
  • No infrastructure modifications required at the airport

Safety Systems:

  • Proprietary hardware + software safety chain
  • Redundant braking paths
  • Certified safety PLCs (Programmable Logic Controllers)
  • Continuous self-diagnostics
  • Controlled, predictable responses to all detected hazards

Fleet Management: EZFleet platform for real-time tracking, route optimization, and mission scheduling

reference airside AV stack autonomous baggage/cargo tug (Underwing Operations) ​

Unique capabilities:

  • Patented sideways drive system for direct lateral movement
  • 360-degree tank turn capability (rotates within its own length)
  • Integrated robotic arms for automatic ULD loading/unloading
  • Electric-powered autonomous operation
  • Designed for close-quarter operations near aircraft on stand

Fleet management: reference airside AV stack fleet integration platform platform for scheduling and monitoring

5. Scaling Plan: From 2 Units to 24 by 2027 ​

Current State (January 2026) ​

  • 2 UISEE tractors in live driverless operations (T1-T4 route)
  • 3 TractEasy EZTow units in operations/trials

Phase 2 (Later 2026) ​

  • 6 additional autonomous tractors on T2-to-aircraft-stands route
  • Operated under CAG-SATS collaboration
  • Supporting baggage operations with potential cargo expansion

Phase 3 (2027) ​

  • Fleet expansion to 24 autonomous tractors total
  • Coverage of baggage, cargo, and equipment towing
  • Multiple routes across T1, T2, T3, T4

What Needs to Change for Scaling ​

Infrastructure Requirements:

  • Extended 5G coverage across all airside areas and terminal connections
  • Autonomous vehicle zone markings painted on airside surfaces (already begun)
  • Clear AV labeling visible to all airside workers
  • Charging infrastructure scaled to support 24+ electric vehicles on continuous duty cycles
  • Potential dedicated AV lanes or traffic management zones

Operational Requirements:

  • Expansion of remote control center capacity (more operators, more screens)
  • Integration with SATS dispatch and flight scheduling systems
  • Coordination protocols between autonomous and manual vehicle traffic
  • Workforce retraining -- CAAS estimates up to 30% of existing airside workforce could experience job redesign over five years
  • Development of new maintenance and troubleshooting capabilities for autonomous fleet

Technology Requirements:

  • Robust fleet management system capable of coordinating 24+ vehicles simultaneously
  • AI-driven dispatch optimization integrated with flight schedules
  • Proven reliability across diverse route types (terminal-to-terminal, terminal-to-aircraft-stand)
  • Demonstrated safety in higher-density mixed traffic scenarios
  • Rain-sensing filtering algorithms validated for Singapore's intense tropical rainfall (up to 50 mm/h)

Strategic Context:

  • Scaling prepares Changi for Terminal 5 operations (opening mid-2030s)
  • T5 will be designed from the ground up to accommodate autonomous operations
  • Experience with T1-T4 deployment informs T5 airside design

6. CAAS Regulatory Approach ​

Advisory Circular AC-139-7-7: Guidance on Use of Autonomous Vehicles at the Airside ​

Effective Date: May 10, 2023

This is the first dedicated regulatory guidance for autonomous vehicles on airport airside in Singapore, issued under the Aviation Navigation Regulations (ANR-139).

Key Regulatory Requirements ​

Safety Risk Assessment:

  • Aerodrome operator (CAG) must conduct a comprehensive safety risk assessment before allowing AV operations
  • Assessment must be revalidated when changes occur to airside driving rules, layout, or operating locations
  • AV operations permitted only if the aerodrome operator is satisfied they can be conducted safely

Vehicle Performance Standards:

  • AVs must not collide with any road or airside user
  • AVs must follow traffic signals, traffic signs, and road markings
  • AVs must adhere to airside speed limits
  • AVs must demonstrate reliable obstacle detection and avoidance

Personnel Requirements:

  • Personnel involved in AV operations (onboard or remote) must have the ability to control or intervene in AV operations at any time
  • Remote operators must maintain continuous monitoring capability
  • Training and certification for remote AV operators required

Operational Approval Process:

  1. Technology vendor demonstrates vehicle capabilities through controlled testing
  2. Progressive testing phases: sterile environment -> live environment with safety driver -> fully driverless with remote supervision
  3. Safety risk assessment submitted to and evaluated by CAAS
  4. Co-funding through CAAS Aviation Development Fund incentivizes responsible innovation
  5. Ongoing monitoring and periodic safety review

Singapore Technical Reference TR68 ​

In addition to the CAAS-specific aviation guidance, autonomous vehicles at Changi must comply with Singapore's national AV standards:

  • TR68 Part 1: Vehicle behavior
  • TR68 Part 2: Vehicle functional safety
  • TR68 Part 3: Cybersecurity (SATS autonomous bus trial specifically assessed against this)
  • TR68 Part 4: Data formats

TR68 was developed under the Singapore Standards Council's Manufacturing Standards Committee as an industry-led effort, with enhanced standards published in September 2021 for safe AV deployment.

CAAS Aviation Development Fund (ADF) ​

CAAS actively co-funds autonomous vehicle initiatives through the ADF, which supports productivity improvement through innovative solutions. This co-funding model:

  • Reduces financial risk for airport operators and ground handlers
  • Encourages multiple technology partners to trial simultaneously
  • Enables progressive development without commercial pressure to deploy prematurely

7. SATS Ground Handling Integration ​

SATS Overview ​

SATS Ltd is Asia's leading provider of gateway services (ground handling and food solutions). At Changi, SATS handles a majority of flights and is the primary ground handler integrating autonomous technology.

Hub Handler of the Future Programme ​

SATS launched its "Hub Handler of the Future" programme to re-engineer ground handling operations for mega airports:

Core Technology Pillars:

  • AI-driven orchestration engine: Under development to leverage live data and predictive analytics for optimizing resource allocation, improving resilience, and raising service quality
  • Automation: Autonomous tractors, forklifts, and airside buses
  • Computer vision: Real-time updates and detection of handling errors to enhance ramp safety
  • Telematics: Vehicle and equipment tracking across the apron

Autonomous Tractor Integration:

  • Autonomous tractors handle the "drive" segment of baggage transport (terminal-to-terminal or terminal-to-aircraft-stand)
  • Human workers focus on "last-mile" operations at aircraft stands: loading/unloading ULDs, positioning containers, and coordination with aircraft crew
  • This division of labor addresses the fact that last-mile operations near aircraft are more complex, variable, and harder to automate than point-to-point driving

Dispatch Integration:

  • Future AI scheduling models will optimize staff and equipment deployment for improved turnaround predictability and on-time performance
  • UISEE's cloud-based intelligent brain provides vehicle dispatch and monitoring
  • EasyMile's EZFleet platform provides real-time tracking and mission scheduling
  • Integration between these fleet management systems and SATS' broader operational planning systems is an active development area

S$250 Million Investment ​

SATS committed over S$250 million to transform Changi operations:

  • S$150 million: Renew and expand GSE fleet over five years (500+ vehicles renewed/refurbished, 100+ new units)
  • S$100 million: Enhance air cargo operations across Changi freight terminals over two years
  • Autonomous technology is a core component of this investment

Workforce Transformation ​

  • SATS is developing multi-skilled career pathways combining technical oversight and ground execution
  • Virtual and augmented reality simulations used for training in operational scenarios
  • Co-developing digital aviation modules, micro-certifications, and AI prompt-engineering training with Institutes of Higher Learning
  • Goal: uplift role value, improve compensation, create new job categories (e.g., remote AV operator, fleet supervisor, autonomous systems technician)

8. Operational Procedures ​

Remote Supervision Model ​

  • Autonomous tractors operate without onboard drivers
  • All operations monitored from a dedicated control center
  • Remote operators supervise via real-time HD video streams (enabled by 5G connectivity)
  • Operators can intervene immediately if the autonomous system encounters a situation it cannot resolve
  • For the TractEasy fleet, some operations are monitored from a tablet interface
  • Former drivers are being retrained to serve as remote AV operators

Typical Mission Flow ​

  1. Dispatch: Mission assigned (currently manual scheduling; future AI-driven dispatch)
  2. Loading: Human workers at the baggage handling area load ULDs onto the tractor's trailer train (up to 4 ULDs)
  3. Transit: Autonomous tractor navigates the 7 km route between terminals (T1-T4) at up to 15 km/h, handling intersections, weather, and other traffic autonomously
  4. Arrival: Tractor arrives at destination baggage area or aircraft stand
  5. Unloading: Human workers unload ULDs for further processing or aircraft loading
  6. Return: Tractor returns empty or picks up outbound baggage for the reverse trip

Safety Protocols ​

  • Autonomous vehicle zone markings painted on airside surfaces
  • Clear AV labels applied to vehicles, visible to all airside workers
  • Emergency stop capability available both on-vehicle and via remote control
  • Redundant braking systems with certified safety PLCs
  • Continuous self-diagnostics with automatic safe-stop on fault detection
  • Geofencing constrains vehicles to approved routes and zones

Charging and Power Management ​

  • Vehicles are electric-powered (lithium-ion or lead-acid battery options)
  • Designed for 24/7 duty cycle operation
  • Automatic charging management integrated into fleet software
  • ACT (Advanced Charging Technologies) provides charging infrastructure for TractEasy deployments
  • Specific charging schedules at Changi have not been publicly disclosed, but the system is designed to minimize downtime through fleet rotation during charging

Maintenance ​

  • Continuous self-diagnostics on all autonomous systems
  • Fleet management platforms provide real-time vehicle health monitoring
  • Sensor calibration and cleaning critical in tropical environment (rain, humidity, heat)
  • LiDAR sensors (Hesai XT series) designed for >30,000 hours service life
  • Maintenance regimes adapted from both aviation GSE standards and automotive autonomous vehicle practices

9. Performance Data ​

Testing Phase Performance ​

  • Test trips: >5,000 during the approximately one-year trial period
  • Distance accumulated: >20,000 km
  • Safety record: Zero accidents / zero safety incidents throughout entire testing phase
  • Weather conditions tested: Day, night, heavy rain, tropical humidity

Operational Performance ​

  • Route distance: 7 km (T1-T4)
  • Payload per trip: Up to 4 ULDs, up to 10 tonnes
  • Maximum speed: 15 km/h (operational speed typically ~10 km/h)
  • Availability: All-weather operations including night and rain

Comparison with Manual Operations ​

Direct quantitative comparison data between autonomous and manual tractor operations at Changi has not been publicly released. However, the following qualitative benefits have been cited:

Efficiency Gains:

  • Autonomous tractors maintain consistent speeds and optimal routing, eliminating human variability
  • Continuous operation capability (no breaks, shift changes, or fatigue)
  • reference airside AV stack trials at other airports demonstrated autonomous tugs could reduce the number of traditional tugs and trailers by two-thirds while substantially cutting carbon emissions

Safety Improvements:

  • Zero accident record over 20,000+ km
  • Elimination of human driving errors (a significant source of airside incidents globally)
  • Consistent adherence to speed limits and traffic rules
  • 360-degree perception eliminates blind spots

Operational Benefits:

  • Frees human workers from repetitive driving tasks to focus on higher-value last-mile operations
  • More reliable baggage handling with smoother aircraft turnarounds
  • Supporting on-time departures and seamless passenger experience
  • Reduced emissions through electric operation

UISEE Global Track Record ​

  • Operational at 21 airports globally without onboard safety operators
  • Completed over 2 million km of real driverless operations (as of early 2026)
  • Full-scenario, true driverless, all-weather operation framework
  • Demonstrated resilience in typhoons, heavy rain, and extreme temperatures

10. Challenges and Lessons Learned ​

Weather: The Defining Challenge ​

Singapore's tropical climate presents arguably the most demanding weather conditions for airside autonomy:

  • Intense rainfall: Singapore regularly experiences rainfall exceeding 50 mm/h. Heavy rain scatters and absorbs LiDAR laser beams, creating distorted signals that compromise obstacle detection accuracy. A dedicated rain-sensing filtering algorithm had to be developed to maintain reliable perception during these events -- this was one of the most significant technical challenges.
  • Humidity and heat: Persistent high humidity and temperatures above 30C stress electronic components, accelerate sensor degradation, and can cause condensation on camera lenses and LiDAR optics.
  • Standing water: Tropical downpours create surface water on the apron that can confuse ground-level sensors and affect traction.

Mixed Traffic Environment ​

The airside environment is fundamentally different from public roads:

  • Unpredictable human behavior: Ground crew walking across vehicle paths, manually driven tractors, pushback tugs, belt loaders, and fuel trucks all share the same space
  • Jet blast avoidance: Determining safe timing to cross behind active aircraft (avoiding jet blast) is a challenge even for experienced human drivers, and remains one of the hardest decision-making problems for autonomous systems
  • Aircraft proximity: Operating near multi-million-dollar aircraft with zero tolerance for contact requires extreme precision and conservative safety margins
  • Foreign Object Debris (FOD): Detection of small objects on the apron is critical -- FOD-related damage costs the industry billions annually, and autonomous vehicles must be able to detect and avoid debris that could be propelled by jet blast

Connectivity Reliability ​

  • Initial operations used 4G and WiFi, which proved insufficient for reliable real-time remote supervision
  • WiFi handover latency spikes between access points caused intermittent connectivity drops for moving vehicles
  • The transition to private 5G addressed these issues but required significant infrastructure investment
  • Maintaining consistent connectivity across the full 7 km T1-T4 route, including any covered sections, required careful 5G network planning

Regulatory Path-Finding ​

  • As the first major airside autonomous vehicle deployment in Singapore, there was no precedent
  • CAAS had to develop AC-139-7-7 guidance alongside the trials, requiring close collaboration between regulator and operators
  • The progressive approach (sterile testing -> live with safety driver -> fully driverless) took several years
  • Compliance with both aviation-specific regulations (ANR-139) and national AV standards (TR68) created a dual regulatory burden

Integration Complexity ​

  • Autonomous tractors must integrate with existing baggage handling systems, flight schedules, and ground handling workflows
  • No single vendor provides the end-to-end solution -- CAG must coordinate between vehicle OEMs, autonomy software providers, 5G network operators, fleet management platforms, and ground handlers
  • Multi-vendor approach (UISEE, EasyMile, reference airside AV stack) provides resilience but adds integration complexity

Workforce Transition ​

  • Retraining drivers as remote operators requires significant investment in training programs
  • Cultural resistance from workers accustomed to traditional driving roles
  • Union engagement critical -- CAAS, CAG, SATS, and their unions collaborate on job redesign programs
  • CAAS estimates up to 30% of the existing airside workforce could experience job redesign over the next five years

Business Case Development ​

  • Industry observers note that the primary challenge is not the technology itself but "finding the right business case and space where this unit can run in real time and at real scale"
  • The high upfront cost of autonomous systems (sensors, compute, 5G infrastructure, control center) requires sustained operational savings to justify
  • Changi's approach of progressive deployment, government co-funding, and multi-vendor competition helps manage this risk

Lessons for the Industry ​

  • No infrastructure modifications required: A key finding -- the TractEasy and UISEE systems both operated on existing airside infrastructure without modifications, significantly reducing deployment barriers
  • Progressive trust-building: The multi-year, multi-phase approach allowed all stakeholders (regulators, operators, workers, airlines) to build confidence gradually
  • Multi-vendor strategy: Testing different autonomous platforms (EasyMile, UISEE, reference airside AV stack) provides comparison data and avoids vendor lock-in
  • 5G is essential: WiFi is insufficient for reliable autonomous vehicle teleoperation on the apron; private 5G is becoming a prerequisite for scaled deployment
  • Tropical climate as proving ground: Systems that work reliably in Singapore's heat, humidity, and intense rain are likely to work anywhere

11. Future Plans ​

Near-Term (2026-2027) ​

Fleet Expansion:

  • 6 additional tractors on T2-to-aircraft-stands route (later 2026)
  • Total fleet of 24 autonomous tractors by 2027
  • Expansion beyond baggage to cargo and equipment towing

Autonomous Bus Trials:

  • SATS, SIA Engineering, and CAG trialing autonomous airside buses for staff transport
  • Two-year program from Q3 2024, co-funded by CAAS
  • Two phases: sterile environment testing, then live operations with safety driver
  • Compliance assessment against TR68 Part 3 (cybersecurity)

reference airside AV stack autonomous baggage/cargo tug Deployment:

  • Phase 2B fleet trials testing coordinated underwing operations for wide-body aircraft turnaround
  • fleet integration platform fleet management platform integration
  • If successful, potential for operational deployment supporting aircraft stand operations

Medium-Term (2027-2030) ​

Additional Vehicle Types:

  • Autonomous cargo dollies (TractEasy has already launched an autonomous cargo dolly product)
  • Autonomous pushback tugs (next logical step in GSE automation)
  • Autonomous fuel bowsers or hydrant dispensers (longer-term potential)
  • Autonomous patrol and inspection vehicles
  • Autonomous firefighting support vehicles (connected to 5G emergency services work)

AI-Driven Operations:

  • SATS AI orchestration engine for automated dispatch and resource optimization
  • Predictive analytics for turnaround time optimization
  • Computer vision for real-time ramp safety monitoring
  • Integration of autonomous vehicle data into airport digital twin

Long-Term (2030s): Terminal 5 and Beyond ​

Terminal 5 Integration:

  • Changi Terminal 5 (opening mid-2030s) will be designed with autonomous operations as a foundational capability
  • Airside layout, traffic management, and infrastructure will be optimized for mixed autonomous/manual vehicle operations
  • Lessons learned from T1-T4 deployments directly inform T5 design

Full Airside Autonomy Vision:

  • Autonomous vehicles shuttling passengers and staff across the airside
  • Majority of point-to-point ground transport automated
  • Human workers focused exclusively on complex last-mile tasks requiring judgment, dexterity, and coordination
  • Autonomous systems integrated into a unified airport operations platform

Replication at Other Airports ​

UISEE's Global Expansion:

  • Already operational at 21 airports globally (primarily in China), including Hong Kong International Airport (70+ autonomous vehicles), Guangzhou Baiyun, Changsha Huanghua, Urumqi, and Beijing Daxing
  • Changi serves as the reference deployment for non-Chinese international airports
  • UISEE claims to be "the world's only provider of sustainable, large-scale commercial L4 autonomous driving solutions for airports"

TractEasy/EasyMile Deployments:

  • Toulouse-Blagnac Airport (France): Fully driverless extended route operations
  • Greenville-Spartanburg International Airport (USA): EZTow deployment
  • Kansai International Airport (Japan): Equipment towing trials
  • Multiple automotive manufacturing plants (e.g., BMW in Germany)

reference airside AV stack Expansion:

  • Changi demonstrations showcase technology to visiting airports and stakeholders
  • Partnership model designed for global replication

Nokia/Private 5G Airport Expansion:

  • Nokia private 5G deployed at Brussels, Vienna, San Sebastian, Miami, Los Angeles, and Paderborn Lippstadt airports
  • Airport 5G for autonomous vehicle support becoming an industry standard
  • Nokia-NTT DATA partnership expanding private 5G globally

Summary Assessment ​

Singapore Changi Airport's autonomous GSE program represents the most methodical and comprehensive approach to airside autonomy at any international airport. Key distinguishing factors:

  1. Multi-vendor, multi-vehicle strategy: Unlike single-vendor deployments, Changi simultaneously evaluates UISEE, EasyMile/TractEasy, and reference airside AV stack, building competitive dynamics and avoiding lock-in.

  2. Regulatory leadership: CAAS issued the first airside-specific autonomous vehicle guidance (AC-139-7-7) and actively co-funds development through the Aviation Development Fund.

  3. Infrastructure investment: The combination of Nokia private 5G and the Singtel 5G Aviation Testbed provides the connectivity foundation that WiFi cannot deliver.

  4. Integration with ground handler transformation: SATS' Hub Handler of the Future programme and S$250M investment ensure autonomous technology is embedded in operational transformation, not deployed as a standalone novelty.

  5. Terminal 5 strategic alignment: Every lesson from the current deployment feeds directly into the design of T5, ensuring next-generation infrastructure is autonomy-native.

  6. Proven safety: 20,000+ km of accident-free operation, validated across Singapore's demanding tropical conditions, establishes confidence for scaling.

The program's deliberate pace -- six years from first controlled testing (2020) to fully driverless operations (2026) -- reflects the safety-critical nature of airport airside operations. The target of 24 vehicles by 2027 will transform Changi from a proof-of-concept to a production-scale autonomous airside operation, setting the standard for the global aviation industry.

Sources ​

Public research notes collected from public sources.

Copyright © 2026 kvynlim