Autonomous Last Mile Delivery Market Size & Share 2026-2035

Autonomous Last Mile Delivery Market Size

The market was valued at USD 1.6 billion in 2026 and is projected to reach USD 11.5 billion by 2035, reflecting a ~24.5% CAGR.

The economic case is rooted in the disproportionate cost of the final leg. The World Economic Forum reported that last-mile activity represented 53% of total shipping cost in 2023, compared with 41% in 2018.^[1] Parcel throughput continues to widen the addressable workload: global volumes reached 161 billion shipments in 2022 and were forecast to reach about 225 billion by 2028. Autonomous systems do not remove every operating cost-remote supervision, maintenance, mapping, and compliance remain material-but they can replace a labor-intensive trip with fleet capacity whose utilization can extend beyond a driver shift. Capgemini identifies last-mile delivery as the most expensive supply-chain component, while U.S. delivery operations remain exposed to driver labor costs.

Near-term adoption will remain shaped less by a uniform global technology curve than by route-specific permissioning. In the United States, the FAA's proposed Part 108 framework seeks to replace the waiver-by-waiver approach for beyond-visual-line-of-sight (BVLOS) operations.^[2] In China, Beijing's regulation effective April 2025 created a city-wide basis for Level 3 and above autonomous-vehicle testing, demonstration, and commercial operation. Such changes determine where fleets can generate delivery density, and thus whether hardware investment turns into recurring revenue.

The early advantage belongs to operators that combine a defined operating domain with an established commercial channel. Walmart-Wing service areas, hospital-linked drone routes, and campus robot networks each concentrate demand before attempting broad geographic coverage. That approach lowers the burden of empty repositioning and operational exception handling. Conversely, a platform with capable hardware but no route density or operating permission is likely to face a longer path to payback.

Key Drivers

E-commerce growth

Parcel growth creates demand for delivery capacity precisely where conventional networks are most costly. Global volumes of 161 billion parcels in 2022, with the 225-billion projection for 2028, increase the value of routes that can be standardized and repeatedly served.^[3] In the U.S., Amazon Logistics processed billions of parcels in 2024, illustrating why large marketplace operators are developing delivery capability as an extension of fulfillment rather than a detached technology experiment.

Fast-turn grocery and restaurant orders strengthen the case because their service promise has little tolerance for batch consolidation. Wing and Walmart reported average delivery times below 19 minutes in their Dallas-Fort Worth service area before announcing a broader U.S. store expansion. For these orders, autonomy is commercially relevant when it improves delivery frequency in a tightly defined zone; a drone or robot cannot compensate for weak inventory placement or an unprofitable order basket.

Rising labor costs

Last-mile delivery relies on a recurring variable-cost base: driving, loading, doorstep handoff, and exception handling. Capgemini reported an average last-mile delivery cost of USD 10.10 against USD 8.08 recovered from consumers in its surveyed model, highlighting the economic gap that automation is intended to narrow.^[4] BLS occupational data also provide the labor-market context for delivery-truck driving as a major operating input.

Automation shifts, rather than eliminates, this cost structure. Fleet operators still need remote assistance, field maintenance, charging, customer support, and regulatory staff. The most credible savings case therefore arises on repetitive routes where one supervision and support layer can serve many completed deliveries, instead of treating a robot or drone as a one-for-one substitute for a courier.

AI and robotics

Commercial viability increasingly depends on the quality of the autonomy stack rather than on vehicle movement alone. Starship reported more than six million deliveries and operations across 80 locations, giving its systems repeated exposure to curbside, pedestrian, and weather-related edge cases. Serve Robotics reported a 2,000-robot fleet and a 99.8% delivery-completion rate, demonstrating why accumulated operating data and fleet monitoring are competitive assets.

Aerial networks face a related requirement: reliable detect-and-avoid, route coordination, and compliance systems. Matternet's September 2024 Light UAS Operator Certificate from Switzerland's FOCA covers advanced SAIL III operations, including BVLOS flights over populated areas. Regulatory acceptance of these operating capabilities turns software validation into a revenue enabler, because it expands the routes a fleet can serve.

Urbanization

Urban density raises both delivery demand and road-network friction. The United Nations projects that 68% of the global population will live in urban areas by 2050. Sidewalk robots can avoid road congestion on suitable pedestrian infrastructure, while drones can bypass surface congestion where flight authorization exists. Neither platform is universally superior: building access, curb design, airspace constraints, payload, and neighborhood acceptance determine which vehicle class can use density productively.

Sustainability

Urban logistics is also becoming an emissions-management issue. Under current growth trends, the WEF projects that last-mile delivery emissions could account for approximately 13% of cities' total carbon emissions by 2030 and approximately 54% of transport-sector emissions. These are projections, not current measurements. Their relevance is that cities may tighten access or operating requirements before the projected outcome occurs, favoring electric delivery systems where they can maintain service levels.

Zipline states that its autonomous drone operations can reduce delivery emissions by up to 97% versus gas-powered ground vehicles. The comparison is route- and load-dependent, but it illustrates the strongest sustainability case: replacing long, time-sensitive ground journeys with a properly utilized electric aerial network, rather than attaching an emissions claim to every autonomous trip.

Key Restraints

Regulatory barriers

Authorization remains fragmented across airspace, roads, sidewalks, and local jurisdictions. The FAA's August 2025 proposed BVLOS rule would establish a performance-based Part 108 pathway, but proposed rules do not confer operating rights. Until a final framework is implemented, operators must still build safety cases and obtain approvals suited to their aircraft and operating environments.

Other jurisdictions show that enabling legislation does not eliminate implementation work. Beijing's 2025 framework supports commercial-operation applications after testing and safety evaluation. The UK's Automated Vehicles Act 2024 establishes a legal framework for self-driving vehicles, while government plans described commercial pilots without a safety driver from spring 2026. Operators must therefore budget for local compliance, liability allocation, and public-realm engagement alongside technology development.

High initial investment

Autonomous delivery requires front-loaded expenditure in vehicles, charging or launch infrastructure, integration with order systems, mapping, safety operations, and market-specific permissions. Hardware acquisition costs vary substantially by platform class, and fleet-scale economics depend on utilization rather than vehicle ownership alone. Recurring maintenance, remote-operations staffing, connectivity, and fleet-management software can delay a route's payback even after vehicles begin service.

The commercial constraint is especially acute before demand density is established. Maersk notes both strong innovation activity and continuing operational-cost and scalability challenges in last-mile solutions. This favors partnerships that pair a fleet provider with a retailer, marketplace, courier, or health system that can supply orders from launch, rather than relying on a standalone fleet to create demand after deployment

Part 108 could reduce the U.S. approval burden if finalized, yet standardization will not erase the practical disciplines of airspace coordination, maintenance, dispatch, and incident response. Investment cases should consequently distinguish a regulatory announcement from operating authorization and distinguish fleet size from productive route density.

Autonomous Last Mile Delivery Market Segment Analysis

By Platform

Drones are projected to grow from USD 0.77 billion in 2026 to USD 4.92 billion in 2035 at a ~22.8% CAGR. They lead revenue at the start of the forecast period because time-sensitive, lightweight orders can justify dedicated aerial infrastructure. Amazon's MK30 is designed for suburban delivery, and Amazon has described a catalog of more than 50,000 eligible items for its drone service.^[5] Fixed-wing systems address a different problem: Zipline describes a 200 km round-trip capability for medical logistics, where road travel time and reliability can outweigh aircraft-network complexity.

Robots are projected to rise from USD 0.69 billion to USD 5.58 billion at a ~26.2% CAGR. Their advantage lies in short, repeatable trips across campuses, neighborhoods, airports, and commercial sites. JD Logistics reported autonomous-vehicle operations in nearly 30 Chinese cities and more than one million kilometers of open-road operation.^[6] Robots need neither rooftop airspace nor a landing area, but their serviceable geography is constrained by sidewalks, crossings, weather, and local operating rules.

Trucks and vans increase from USD 0.14 billion to USD 1.00 billion at a ~24.5% CAGR. They are relevant where autonomous hub-to-spoke movement supports the last mile rather than completes a doorstep transaction. DHL Supply Chain and Volvo began autonomous-truck operations on Texas freight lanes in December 2024 with a safety driver present. This segment's route economics depend on freight-lane repetition and freight utilization, not on the delivery characteristics that favor a sidewalk robot.

By Delivery Mode

B2B grows from USD 0.55 billion in 2026 to USD 4.34 billion in 2035 at a ~25.8% CAGR, the highest rate among delivery modes. Predictable transfers between warehouses, stores, clinics, laboratories, and local distribution points can support defined service-level agreements and reduce failed-delivery risk. Matternet's Zürich hospital route demonstrates the importance of controlled institutional endpoints in early medical-drone deployment.

B2C remains the largest delivery mode, rising from USD 0.76 billion to USD 5.31 billion at a ~24.2% CAGR. Its value rests on convenience, but its economics are sensitive to order density, customer availability, and the cost of exceptions. C2C grows from USD 0.29 billion to USD 1.85 billion at a ~22.6% CAGR; it is less mature because pickup, identity, returns, and chain-of-custody workflows are harder to standardize when both endpoints are consumers.

By Range

Short-range delivery below 20 km is projected to expand from USD 1.14 billion in 2026 to USD 8.11 billion in 2035 at a ~24.3% CAGR. It fits the operational geometry of sidewalk robots and multi-rotor drones: compact zones, high delivery frequency, and nearby inventory. Long-range delivery above 20 km grows from USD 0.46 billion to USD 3.39 billion at a ~25.0% CAGR, led by medical drone corridors and autonomous freight links. The latter can have a higher service value, but also needs a more robust operating authorization and network design.

By Solutions

Hardware rises from USD 0.90 billion in 2026 to USD 6.34 billion in 2035 at a ~24.1% CAGR, reflecting fleet acquisition, airframes, chassis, sensing, batteries, and charging infrastructure. Software grows faster, from USD 0.42 billion to USD 3.21 billion at a ~25.2% CAGR, because scaled fleets require dispatch, localization, monitoring, route optimization, and compliance records. Services expand from USD 0.27 billion to USD 1.95 billion at a ~24.5% CAGR as maintenance, integration, and operating support become recurring requirements.

The solution mix will evolve with fleet maturity. Early programs purchase hardware and build local operations; established networks derive more value from keeping many assets safely utilized. That creates an opening for delivery-as-a-service and robot-as-a-service models, provided the provider can fund the hardware and retain operational accountability.

By Application

E-commerce is the largest application, increasing from USD 0.68 billion in 2026 to USD 4.84 billion in 2035 at a ~24.4% CAGR. Food and grocery rise from USD 0.24 billion to USD 1.65 billion at a ~23.8% CAGR, where speed and basket frequency matter more than payload. Parcel and courier services grow from USD 0.44 billion to USD 3.24 billion at a ~24.9% CAGR, benefiting from operators' existing address, tracking, and fulfillment systems.

Pharmaceutical delivery advances from USD 0.19 billion to USD 1.36 billion at a ~24.7% CAGR. Its commercial proposition centers on time certainty, documented custody, and access rather than simply lower delivery cost. Zipline has announced Platform 2 relationships with U.S. health systems, including Mayo Clinic, Cleveland Clinic, and Intermountain Health. Furniture and appliance delivery remains smaller, rising from USD 0.06 billion to USD 0.40 billion at a ~24.1% CAGR, because its payload and handling requirements align more closely with autonomous vans and trucks than with current drones or sidewalk robots.

These application patterns are consistent with the broader market structure, where retail/e-commerce remains a major demand driver while healthcare, food, logistics, and other specialized applications create additional opportunities for autonomous delivery platforms.

Software's leading growth rate signals an important margin shift. As fleets expand, the scarce capability becomes safe orchestration across vehicles, orders, exceptions, and permissions. Healthcare can support this investment earlier because the cost of delay can exceed transport cost, whereas food and e-commerce require very high order density to offset support and infrastructure costs.

Autonomous Last Mile Delivery Market Regional Analysis

North America

North America is projected to grow from USD 0.56 billion in 2026 to USD 3.39 billion in 2035 at a ~22.1% CAGR. The U.S. rises from USD 0.50 billion to USD 3.01 billion at a ~22.2% CAGR, and Canada from USD 0.07 billion to USD 0.38 billion at a ~21.2% CAGR. U.S. scale is supported by major retailer and food-platform partnerships: Wing's Walmart expansion and Serve's Uber Eats fleet demonstrate two different routes to demand aggregation. The key swing factor remains implementation of a scalable BVLOS pathway.

Europe

Europe expands from USD 0.48 billion in 2026 to USD 3.23 billion in 2035 at a ~23.5% CAGR. Germany grows from USD 0.14 billion to USD 0.94 billion at a ~23.6% CAGR, while the rest of Europe moves from USD 0.35 billion to USD 2.29 billion. Europe's opportunity rests on a mix of medical-drone corridors, dense urban delivery, and a structured legal environment. Matternet's Swiss authorization provides a concrete example of advanced drone permissioning, while the UK's automated-vehicle framework supports future ground-vehicle pilots.

Asia Pacific

Asia Pacific is the fastest-growing region, from USD 0.40 billion in 2026 to USD 3.69 billion in 2035 at a ~27.9% CAGR. China increases from USD 0.18 billion to USD 1.59 billion at a ~27.6% CAGR, and the rest of Asia Pacific from USD 0.23 billion to USD 2.10 billion at a ~28.0% CAGR. China's combination of delivery volume, dense urban demand, and operating deployments gives it a different starting point from markets still focused on pilots. JD Logistics' reported city coverage and Beijing's city-wide regulatory framework connect operating data to a widening commercial basis.^[7]

India illustrates the distinction between trial progress and market-scale operation. Swiggy participated in BVLOS trials that completed more than 300 food and medicine deliveries in Uttar Pradesh and Punjab, and later pursued an Instamart drone model for B2B middle-mile movement. Zomato acquired TechEagle in 2018, tested a 5 km food-drone route in 2019, and obtained DGCA permission for BVLOS testing in 2020. These are meaningful capability-building steps, but they should not be treated as evidence of unrestricted consumer drone delivery.

Latin America

Latin America grows from USD 0.08 billion in 2026 to USD 0.71 billion in 2035 at a ~26.7% CAGR. Brazil rises from USD 0.03 billion to USD 0.24 billion at a ~26.8% CAGR, and the rest of Latin America from USD 0.06 billion to USD 0.47 billion. Large urban areas and dispersed delivery demand create a potential use case, but operating scale will depend on national and municipal rules, reliable charging and communications infrastructure, and commercially viable local partnerships.

Middle East & Africa

The Middle East & Africa rises from USD 0.06 billion in 2026 to USD 0.48 billion in 2035 at a ~25.0% CAGR. The UAE grows from USD 0.02 billion to USD 0.14 billion at a ~25.2% CAGR, while the rest of the region moves from USD 0.05 billion to USD 0.34 billion. The region's opportunity is bifurcated: Gulf markets can support planned urban logistics pilots, while health and essential-goods logistics offer a clearer case in areas where road access is less reliable. Zipline's African medical-delivery experience shows why the latter can be a more durable early application than discretionary consumer delivery.

The most transferable cross-regional model is institutional and corridor-based deployment. Hospital systems, campuses, retail clusters, and hub-to-store transfers reduce the number of uncontrolled variables at launch. Consumer doorstep delivery can scale after these networks establish safety credibility and utilization; it is not necessarily the first economically defensible use case in every region.

Autonomous Last Mile Delivery Market Share & Competitive Landscape

The 2025 market-share structure is led by Amazon at 15.3%, followed by Starship Technologies at 9.6%, Nuro at 6.2%, JD.com at ~4.5%, Flirtey at ~3.3%, Kiwibot at ~3.2%, UPS at ~1.3%, Matternet at ~1.2%, Zipline at ~0.3%, and Wing Aviation at ~0.2%. These proprietary estimates reflect different business models: integrated retailers and couriers, fleet operators, platform providers, and technology suppliers cannot be compared solely by vehicle count.

Amazon is integrating Prime Air into a large fulfillment ecosystem, while DHL is pursuing multimodal automation, including autonomous truck operations in Texas.^[8] Flirtey has partnered with DoorDash and Uber for U.S. drone-delivery integration. JD.com and JD Logistics combine e-commerce demand with autonomous ground-logistics deployment. Matternet focuses on regulated medical-drone networks, and UPS Flight Forward has used Matternet in healthcare delivery. Nuro has shifted toward licensing its Nuro Driver technology, demonstrating that autonomy developers may monetize software and vehicle platforms without operating every route themselves.^[9]

Starship, Wing, and Zipline illustrate three different scaling logics: dense ground-robot networks, retailer-linked aerial delivery, and medical logistics. Manna Aero operates European drone delivery, while Ottonomy.IO's Ottobot has been deployed in a DPD UK autonomous "locker on wheels" program. Serve Robotics scales sidewalk delivery through Uber Eats, and Postmates, as part of Uber, is connected to that ecosystem. Kiwibot has developed robot deployments across multiple internationalLast Mile Delivery markets.

Regional and emerging competitors broaden the competitive base. RoboSense is an HKEX-listed LiDAR and perception-system supplier to automotive and robotics ecosystems; it reported approximately 260,000 LiDAR units sold in 2023 and should be viewed as a component and perception supplier, not a delivery operator. Segway Robotics supplied 1,000 sidewalk robots to Coco for U.S. deployment and has supported grocery and restaurant robot service in Spain through Goggo Network. SF Express combines courier distribution with autonomous programs, including an order for 100 ES1000 uncrewed cargo aircraft and a driverless freight pilot with Rino.ai.

TeleRetail Robotics, now Aitonomi AG, develops outdoor delivery robots and has received ESA support for feasibility work with Swiss Post; its Aito platform also featured in a Coca-Cola European Partners London pilot. Yandex Delivery's Rover program reached U.S. campuses through a Grubhub partnership and reported more than 300,000 cumulative orders by late 2023. Refraction AI, a University of Michigan spinout, raised a USD 4.2 million seed round in 2021 and piloted approximately 10 REV-1 bike-lane robots with Chick-fil-A in Austin in 2022. No unsupported later funding figure is retained.

Competitive advantage is likely to accrue through authorization, route density, operating data, and customer access in combination. A perception supplier such as RoboSense benefits as fleets are built, but it does not control the delivery relationship. A retailer or courier can control demand but may depend on a specialized operator for aircraft, robots, and safety systems. The most defensible positions align these capabilities without assuming that a widely announced partnership has already achieved scaled operations.

Recent Industry Developments

Matternet's Swiss certification - September 2024. Matternet received a Light UAS Operator Certificate from Switzerland's FOCA for advanced SAIL III operations, including BVLOS flights over populated areas.

Beijing autonomous-vehicle regulation - April 2025. Beijing's city-wide regulation took effect, covering Level 3 and above autonomous-vehicle testing, demonstration, and commercial-operation pathways.

Wing and Walmart expansion - June 2025. Wing announced plans to add drone delivery from 100 Walmart stores across five U.S. cities after service in the Dallas-Fort Worth area.

Flytrex platform partnerships - June-September 2025. DoorDash launched Flytrex drone delivery in Dallas-Fort Worth in June, and Uber later announced an investment and planned integration with Flytrex.

FAA proposed BVLOS rule - August 2025. The FAA proposed a performance-based Part 108 framework for certain BVLOS drone operations, a key regulatory development for U.S. package delivery.

Serve Robotics fleet milestone - December 2025. Serve reported completion of a 2,000-robot autonomous delivery fleet for its Uber Eats-linked operations.