Drone delivery vs autonomous ground vehicles: cost, speed, and sustainability compared

Why It Matters

Last-mile delivery accounts for roughly 53 percent of total shipping costs and generates an estimated 30 percent of urban transport emissions (World Economic Forum, 2024). With global e-commerce parcel volumes projected to exceed 256 billion by 2027 (Pitney Bowes, 2025), logistics operators face intense pressure to cut both costs and carbon simultaneously. Two technologies have emerged as leading contenders to replace conventional van-based delivery: unmanned aerial vehicles (drones) and autonomous ground vehicles (AGVs), sometimes called sidewalk delivery robots. Each approach offers distinct advantages in speed, payload capacity, energy consumption, and regulatory readiness. Yet choosing between them requires more than headline metrics. Terrain, population density, package weight distribution, and local regulations all shape the total cost of ownership and environmental footprint. This comparison guide provides a structured, evidence-based framework to help supply chain leaders evaluate both options.

Key Concepts

Drone delivery refers to the use of small unmanned aircraft systems (sUAS), typically electric multirotor or fixed-wing VTOL platforms, to transport packages directly from a depot or micro-fulfillment center to a customer location. Most commercial platforms carry payloads of 2 to 5 kilograms over distances of 5 to 25 kilometers. Wing (Alphabet) completed more than 350,000 commercial deliveries across Australia, Finland, and the United States by mid-2025 (Wing, 2025).

Autonomous ground vehicles are electric, self-driving robots that navigate sidewalks, bike lanes, or roadways at low speeds (typically 6 to 25 km/h). They carry payloads ranging from 10 to 200 kilograms depending on vehicle class. Starship Technologies reported surpassing 7 million autonomous deliveries across 20 countries by late 2025 (Starship Technologies, 2025), making it the highest-volume operator in this category.

Last-mile delivery is the final leg of the supply chain from a local distribution hub to the end customer. It is the most expensive and emissions-intensive segment because of stop-and-go routing, failed delivery attempts, and low vehicle utilization.

Carbon intensity per delivery measures the grams of CO₂ equivalent emitted per package delivered. Conventional diesel vans typically emit 300 to 600 gCO₂e per stop depending on route density (McKinsey, 2024). Both drones and AGVs offer significant reductions because they are electrically powered and right-sized for small parcels.

Head-to-Head Comparison

Sources: Wing (2025), Nuro (2025), McKinsey (2024), FAA (2025).

Both modalities produce dramatically lower emissions than diesel vans. A University of Michigan life-cycle analysis found that drone delivery reduces greenhouse gas emissions by 54 to 84 percent relative to conventional truck delivery for lightweight packages under 3 kg (Stolaroff et al., updated 2024). Ground robots show similar advantages for heavier multi-item orders because their larger battery capacity is offset by multi-drop efficiency.

Cost Analysis

Drone delivery cost per package has fallen sharply as fleet scale increases. Wing reported an average cost of $2.50 to $4.00 per delivery at commercial scale in suburban Australia, down from over $10 per delivery during early trials (Wing, 2025). Zipline, which operates the largest drone logistics network globally with over 1 billion kilometers flown, achieves costs below $2.00 per delivery for medical supplies in Rwanda and Ghana where road infrastructure is poor (Zipline, 2025). Key cost drivers include aircraft depreciation, battery replacement cycles, remote pilot staffing ratios, and ground infrastructure (launch pads, charging stations).

AGV delivery cost per package benefits from lower capital cost per unit and multi-drop capacity. Starship Technologies reported costs of $1.50 to $3.00 per delivery in dense university campus and suburban environments (Starship Technologies, 2025). Nuro, which operates larger road-going autonomous vehicles, targets costs below $5.00 per delivery for grocery-sized loads (Nuro, 2025). Key cost drivers include fleet size, teleoperation staffing, maintenance, and mapping/localization updates.

Conventional van delivery costs $7 to $12 per package in urban environments (Deloitte, 2024), providing a clear economic incentive for automation.

Total cost of ownership (TCO) considerations extend beyond per-delivery costs. Drones require aviation insurance, airspace management systems, and compliance with evolving beyond-visual-line-of-sight (BVLOS) regulations. AGVs need sidewalk permits, geofenced operational design domains, and pedestrian safety validation. Both require micro-fulfillment centers positioned closer to customers than traditional warehouses, adding real estate costs but reducing transit distances.

Use Cases and Best Fit

Drones excel in:

• Rural and semi-rural delivery where road networks are sparse, distances are long, and customer density is low. Zipline's operations in East Africa and expansion into the U.S. demonstrate that drones can reach locations in 15 minutes that would take vans over an hour.
• Urgent and time-critical shipments such as medical supplies, lab samples, and emergency parts. Zipline delivers blood products and vaccines to over 4,000 health facilities across seven countries (Zipline, 2025).
• Suburban residential delivery of lightweight e-commerce packages. Wing's operations in Logan, Australia, and the Dallas-Fort Worth metroplex show consistent sub-15-minute delivery for food and pharmacy items.

AGVs excel in:

• Dense urban and campus environments where short distances, flat terrain, and high order density favor multi-drop routes. Starship operates on over 30 university campuses and in suburban neighborhoods across the U.S. and Europe.
• Grocery and multi-item orders where payload requirements exceed drone capacity. Nuro's R3 vehicle carries up to 230 kg of goods for partners including Kroger, Domino's, and FedEx.
• Consistent all-weather operation because ground robots are less affected by wind, rain, and temperature extremes than aerial platforms.

Hybrid deployments are increasingly common. Amazon has tested both its MK30 drone platform and Scout ground robot to serve different delivery windows and package profiles from the same fulfillment center (Amazon, 2025). This hybrid approach allows operators to route lightweight urgent orders via drone and heavier multi-item baskets via AGV, optimizing cost and speed across the full order mix.

Decision Framework

Supply chain leaders evaluating these technologies should consider the following criteria:

1. Package weight distribution. If more than 70 percent of orders weigh under 3 kg, drones become cost-competitive. If the median order exceeds 5 kg, AGVs or hybrid models are preferable.

2. Service area geography. Map the delivery radius, terrain type, and road quality. Drones bypass infrastructure constraints entirely, making them superior for areas with poor roads, water crossings, or mountainous terrain. AGVs require paved, accessible routes.

3. Delivery speed requirements. For sub-30-minute delivery promises, drones offer a structural advantage. AGVs are better suited to scheduled delivery windows of one to three hours.

4. Regulatory environment. Check local drone regulations including BVLOS authorization status, airspace restrictions near airports, and noise ordinances. AGV regulations vary by municipality but are generally less restrictive. The FAA expanded Part 135 drone delivery certificates to 14 operators by early 2026, while AGV operators navigate a patchwork of state and local permits (FAA, 2025).

5. Order volume and density. High-density routes favor AGVs because of multi-drop capability. Scattered, low-density delivery points favor drones.

6. Sustainability targets. Both platforms reduce emissions by 50 to 80 percent versus diesel vans. However, drones powered by renewable electricity in low-carbon grids achieve the lowest per-delivery emissions. Evaluate the local grid carbon intensity and the availability of on-site solar charging.

Key Players

Established Leaders

• Wing (Alphabet) — Largest commercial drone delivery operator by geographic reach, with over 350,000 deliveries completed across three continents.
• Zipline — Operates the world's largest drone logistics network, serving healthcare and e-commerce across seven countries with over 1 billion flight kilometers.
• Starship Technologies — Market leader in sidewalk delivery robots with 7 million completed deliveries across 20 countries.
• Nuro — Road-going autonomous delivery vehicle operator partnering with Kroger, Domino's, and FedEx in the United States.

Emerging Startups

• Manna Aero — Irish drone delivery company operating suburban food delivery in Ireland and expanding into the U.S. market.
• Serve Robotics — Sidewalk delivery robots operating in Los Angeles through partnerships with Uber Eats.
• Flytrex — Drone delivery operator focused on suburban U.S. markets with FAA Part 135 approval.
• Coco — Remote-operated sidewalk delivery robots serving restaurants in Los Angeles and Austin.

Key Investors/Funders

• Breakthrough Energy Ventures — Invested in Zipline and multiple autonomous logistics startups focused on decarbonization.
• SoftBank Vision Fund — Major backer of Nuro and other autonomous vehicle companies.
• a16z (Andreessen Horowitz) — Investor in Serve Robotics and multiple last-mile automation platforms.
• Toyota Ventures — Invested in autonomous delivery and logistics robotics companies.

FAQ

How do drone and AGV delivery compare on carbon emissions? Both technologies reduce last-mile emissions by 50 to 84 percent compared to diesel vans, according to life-cycle analyses by the University of Michigan (Stolaroff et al., 2024) and McKinsey (2024). Drones tend to have a slight edge for lightweight single-item deliveries because of shorter trip times, while AGVs can be more efficient per package on multi-drop routes. The electricity source matters significantly: operations powered by renewable energy can achieve near-zero operational emissions.

Are drones or ground robots safer for urban environments? Both technologies have strong safety records. Starship reports zero serious incidents across 7 million deliveries. Drone operators like Wing and Zipline have completed hundreds of thousands of flights with no injuries to bystanders. However, drones face greater public concern about overhead flight, noise, and privacy. Ground robots face challenges with pedestrian interactions, accessibility on sidewalks, and curb navigation. Regulatory bodies require extensive safety validation for both.

What regulatory approvals are needed? Drone delivery requires an air operator certificate (Part 135 in the U.S.), BVLOS waivers or approvals, and compliance with airspace restrictions. The FAA, EASA, and CASA have all expanded BVLOS pathways since 2024. AGVs need local sidewalk or road-use permits, which vary widely by jurisdiction. In the U.S., over 20 states have enacted legislation permitting personal delivery devices on sidewalks, but requirements for speed limits, weight restrictions, and insurance differ.

Can these technologies work together in a single logistics network? Yes. Hybrid deployments are gaining traction. Amazon, Walmart, and several logistics startups are testing integrated networks where drones handle urgent, lightweight deliveries and ground robots serve heavier, multi-item orders from the same fulfillment nodes. This approach maximizes asset utilization and allows operators to match the right vehicle to each order profile.

What is the timeline for widespread adoption? McKinsey (2024) projects that autonomous last-mile delivery (drones and AGVs combined) will handle 10 to 25 percent of last-mile volumes in major markets by 2030. Zipline and Wing are already operating at commercial scale in select geographies. Scaling depends on regulatory expansion of BVLOS operations, public acceptance, and continued cost reduction in vehicle hardware and batteries.

Sources

• World Economic Forum. (2024). The Future of the Last-Mile Ecosystem. WEF.
• Pitney Bowes. (2025). Parcel Shipping Index: Global E-Commerce Volume Projections. Pitney Bowes.
• McKinsey & Company. (2024). Autonomous Delivery: The Next Frontier in Last-Mile Logistics. McKinsey.
• Stolaroff, J. et al. (2024). Energy Use and Life Cycle Greenhouse Gas Emissions of Drones for Commercial Package Delivery (Updated Analysis). University of Michigan / Nature Communications.
• Wing. (2025). Wing Commercial Delivery Milestones and Cost Analysis. Alphabet / Wing.
• Zipline. (2025). Zipline Annual Impact Report: 1 Billion Kilometers Flown. Zipline International.
• Starship Technologies. (2025). 7 Million Deliveries: Fleet Performance and Safety Report. Starship Technologies.
• Nuro. (2025). R3 Platform: Autonomous Delivery Vehicle Specifications and Partner Results. Nuro.
• Deloitte. (2024). The Last Mile: Urban Delivery Economics and Decarbonization Opportunities. Deloitte Insights.
• FAA. (2025). Part 135 Drone Delivery Certificates and BVLOS Authorization Status. Federal Aviation Administration.
• Amazon. (2025). Prime Air and Last-Mile Automation: Hybrid Delivery Network Update. Amazon.