Broker: Maybank Research Pte Ltd
Date of Report: July 29, 2025

Grab Leads the Autonomous Vehicle Revolution: In-Depth Analysis of AV Economics, Partnerships, and Financial Impact

Introduction: Grab’s Strategic Position in the Autonomous Vehicle Shift

Grab Holdings is strategically poised to capitalize on the rise of autonomous vehicles (AVs) in Southeast Asia, particularly in Singapore. Early investments and partnerships in AV technology, coupled with a robust balance sheet and operational expertise, position Grab to benefit significantly as AV adoption accelerates. This comprehensive analysis explores the evolving AV landscape, competitive economics, and the financial implications for Grab and its peers.

Key Companies Analyzed

• Grab Holdings (GRAB US)
• Baidu (BIDU US)
• Tesla (TSLA US)
• Waymo (Alphabet/GOOG US, Unlisted)
• Didi (DIDIY US)
• Uber (UBER US)
• Mobileye (MBLY US)
• Lyft (LYFT US)
• ComfortDelGro (CD SP)

Grab’s Competitive Edge in Singapore’s AV Transition

Grab is set to lead Singapore’s AV transition, with a blue-sky scenario projecting AV operational costs to fall to USD 0.52/mile by 2030, compared to rising driver costs of USD 0.85/mile (assuming 4% annual inflation). If 20% of Grab’s Singapore fleet transitions to AVs, this could yield annual savings of approximately USD 71 million and boost Grab’s net present value by 7%. Early AV partnerships with A2Z, Motional, and WeRide, along with Grab Rentals’ established fleet management infrastructure, enable Grab to play both a platform and fleet operator role. Its strong balance sheet further supports investment and reduces the risk from new technology disruptors.

AV Partnerships: The New Path to Scale

The AV ride-hailing landscape is shifting from asset-light models to infrastructure-heavy operations, requiring significant investment in autonomy tech, platform integration, and fleet management. Global trends reveal that partnerships, not rivalry, are driving AV scale. Notable examples include:

• Waymo’s integration with Uber in the US
• Mobileye’s partnership with Lyft
• Baidu’s vertically integrated Apollo Go in China, which has completed over 11 million rides, though it still trails Didi’s 33 million daily trips
• South Korea and Middle East: Hyundai, WeRide, and Baidu forming alliances with local ride-hailing apps and governments

AV Unit Economics: The Race to Cost Efficiency

Autonomous vehicle cost per mile is nearing competitiveness with human-driven alternatives:

• Waymo’s operational cost: USD 2.31/mile (excluding platform overheads and R&D)
• Uber’s human-driven cost: USD 1.97/mile
• Amortized R&D can push Waymo’s cost above USD 3.3/mile
• Vision-led systems (Tesla FSD, Baidu Apollo Go) are narrowing the gap, with Apollo Go’s RT6 priced under USD 30,000 compared to Waymo’s >USD 100,000 LiDAR-reliant AV

With AI-driven vision technology and declining bill-of-materials, AV cost per mile is expected to fall below human-driven models, especially in mature markets like Singapore, where AV tech stack cost per mile is already estimated at USD 0.84 at full utilization, versus driver earnings of USD 0.70.

Challenges in Emerging ASEAN: AVs vs. Cheap Labor

In emerging ASEAN markets such as Indonesia and Vietnam, AV adoption faces significant hurdles:

• Driver costs are just USD 0.2-0.3/mile, far lower than AV operational costs of USD 0.7-0.8/mile
• Poor road infrastructure and informal traffic patterns further increase localization costs and complexity
• Political sensitivity around driver displacement makes mass deployment in the medium term unlikely

Initial AV rollouts are likely to be limited to controlled zones and specialized use cases, with wide-scale adoption still distant.

Global Case Studies: Partnerships Dominate AV Rollouts

United States:

• Waymo, Cruise, and Aurora lead, but capital intensity and operational complexity favor partnerships
• Waymo’s robotaxis are integrated into Uber’s app in Austin and Atlanta; Mobileye partners with Lyft

China:

• Baidu pursues a full-stack, fully owned AV approach with Apollo Go, handling software, hardware, mapping, and fleet
• Apollo Go: 11+ million robotaxi rides, but still a minor player compared to Didi’s 33 million daily trips
• Didi is ramping up AV efforts, collaborating with GAC Aion to produce Level 4 robotaxis and investing in its own autonomy system

South Korea:

• Hyundai’s Level 4 RoboRide pilot operates in Seoul via Jin Mobility’s i.M app; Kakao Mobility in talks with Waymo and Baidu

Middle East:

• WeRide and Uber are piloting robotaxi trials in Abu Dhabi, aiming for full driverless service by 2026
• Pony.ai and Baidu’s Apollo Go are entering Dubai and Saudi Arabia, with ambitions for rapid expansion

Traditional vs. AV Ride-Hailing: Business Model Shift

Traditional ride-hailing is asset-light, with drivers owning vehicles and minimal capital expenditure for the platform. AV ride-hailing, by contrast, is capital-intensive, requiring:

• Ownership or leasing of autonomous vehicles
• Investment in depots, charging infrastructure, and maintenance
• High up-front costs for sensor hardware, compute, and R&D
• Complex fleet management and operational readiness

A hybrid model is emerging, where AV technology providers partner with ride-hailing platforms for user access, demand aggregation, and localized operational expertise.

AV Economics: Current State and Future Trajectory

A detailed breakdown of current US unit economics:

Waymo’s heavy R&D investments (estimated \$25-30 billion) mean its amortized cost per mile is about \$3.31. Platform-related costs (marketing, incentives, customer support) are not included and can further impact profitability.

Vision-Led Autonomy: The Future of Cost-Effective AVs

Vision-based AV systems, such as Tesla’s Full Self-Driving (FSD) and Baidu’s Apollo Go, offer a path to lower costs and greater scalability by relying on camera arrays, occupancy networks, and AI-driven perception—dramatically reducing sensor suite costs compared to LiDAR/radar-heavy approaches.

AV Operating Model: The Three-Stack Approach

Scalable AV ride-hailing requires three specialized layers:

1. AV Technology Stack: Led by players like Waymo, Tesla, Baidu, focusing on perception, planning, and vehicle integration. They are streamlining tech platforms and outsourcing non-core activities.
2. Ride-Hailing Platform: Uber, Grab, Didi provide demand aggregation, dynamic pricing, and logistics, filling critical gaps in predictive demand modeling and fleet balancing.
3. Fleet Operations: Infrastructure partners ensure fleet uptime with charging, cleaning, repairs, and depot management, acting as essential “fleet anchors.”

Singapore Spotlight: AV Economics at Full Utilization

Singapore provides a unique testbed for AV economics:

• AV hardware cost per car: USD 80,000, translating to USD 0.4/mile (based on 212 km/day, 10-year amortization)
• Estimated fleet operations and maintenance: USD 0.34/mile
• Foundational cost amortization: USD 0.3/mile
• Total AV cost per mile: USD 0.84
• Driver cost per mile: USD 0.70 (SGD 0.5/km, monthly net earnings SGD 3,000)

As AV hardware and foundational costs decline, AV cost per mile is expected to undercut human-driven rides. However, this requires high utilization rates, which may be challenging to achieve in reality.

Emerging ASEAN: Structural Barriers to AV Adoption

• Driver earnings in Indonesia, Vietnam, and Philippines: USD 0.2-0.3/mile
• AV costs: USD 0.7-0.8/mile
• Fragmented infrastructure, informal traffic patterns, and high localization costs slow deployment
• Labor displacement is a sensitive political and social issue
• Initial AV deployments likely limited to airports, business parks, and night shuttles

Grab’s AV Initiatives: Platform and Fleet Operator Potential

Grab has launched several AV initiatives in Singapore:

• MOUs with Autonomous A2Z, Motional, WeRide, and Zelos for AV applications in mobility and delivery
• Private autonomous shuttle service for employees between one-north HQ and MRT station; equipped with LiDAR, radar, cameras, and safety drivers
• Food delivery robot pilot on Sentosa beach using Neolix robots and AI fleet platform

Grab’s dual strength in demand aggregation and fleet management (via Grab Rentals) positions it to orchestrate both the ride-hailing platform and AV fleet operations. Its hybrid fleet model will deploy AVs for predictable, high-utilization routes (e.g., airport transfers, business district commutes), while human drivers will handle unpredictable peak and surge demand.

Financial Impact for Grab: Blue-Sky Scenario

Projected cost evolution (2030 vs. 2025):

• AV costs per mile expected to drop from USD 0.84 to USD 0.52 (vision-based autonomy drives hardware cost down from USD 80,000 to USD 30,000)
• Driver cost per mile expected to rise from USD 0.70 to USD 0.85 (4% annual inflation)
• If 20% of Grab’s Singapore fleet transitions to AVs, annual pre-tax savings could reach USD 86 million, with USD 71 million post-tax
• NPV uplift estimated at 7% (8% WACC, 5% terminal growth)

Further upside could come from AV adoption in markets like Kuala Lumpur and AV tech in Grab’s Deliveries segment.

ComfortDelGro: Preparing for the Autonomous Future

ComfortDelGro is actively preparing its workforce for AV adoption:

• Launched training and capability development for AV safety operators in partnership with Moovita
• Established a SGD 30 million AV Centre of Excellence for AV operations, maintenance, and technology development
• Pilot program with Pony.ai in China: 100 robotaxis in Guangzhou (March 2025 onwards)
• Plans to expand training for remote operators, maintenance specialists, data analysts, and fleet management roles
• AV tech expected to help address driver shortages and extend transport to under-served areas

Pricing of self-driving services remains a challenge, balancing passenger affordability while avoiding undercutting human-driven taxis.

ESG and Sustainability Highlights

Grab’s ESG initiatives are robust:

• 349,986 tonnes of GHG emissions avoided in 2023
• 6.3% of all distance travelled on low/zero emission modes (EVs, hybrids, cyclists, walkers)
• WWF-Singapore Plastic Action Pact signatory; eco-friendly packaging initiatives
• KPIs and disclosures on emissions, resource usage, and workforce diversity
• Targets: Zero packaging waste and carbon neutrality by 2040, 100% renewable energy by 2030, 40% women in leadership by 2030

Grab’s ESG score is above average at 46, with strong qualitative metrics.

Grab Financial Summary and Outlook

• Adjusted EBITDA projected at USD 485 million for FY25E, reaching USD 1.05 billion by FY27E
• 2024-2027E on-demand GMV CAGR of 16%; adjusted net revenue CAGR of 20%
• Take-rates expected to remain stable

Conclusion: Opportunities Outweigh Risks for Grab

Grab stands out as a net beneficiary in the AV revolution. Its early investments, operational expertise, and strong balance sheet enable it to capture upside from AV adoption while remaining resilient to slower-than-expected timelines. Even in a scenario where AV progress is delayed, Grab’s core human-driven platform ensures limited downside risk. With the industry pivoting toward hybrid AV models and partnerships, Grab is structurally positioned to lead Southeast Asia’s mobility transformation.