Explainer: Logistics automation, drones and last-mile delivery

Why It Matters

Last-mile delivery accounts for roughly 53 percent of total shipping costs and generates up to 30 percent of urban transport emissions, according to the World Economic Forum (2024). With global parcel volumes projected to exceed 256 billion by 2027 (Pitney Bowes, 2025), the environmental footprint of moving goods from distribution centres to doorsteps is intensifying at a pace that conventional diesel fleets cannot sustainably support. Logistics automation, drone delivery, and autonomous last-mile solutions represent a convergence of technologies that can slash carbon intensity per parcel by 40 to 80 percent while simultaneously reducing delivery costs by 20 to 40 percent (McKinsey, 2025). For sustainability professionals, understanding these technologies is no longer optional: regulatory pressure from the EU Corporate Sustainability Due Diligence Directive, Scope 3 reporting mandates under CSRD, and consumer demand for greener delivery options are accelerating adoption timelines across every major market.

Key Concepts

Logistics automation encompasses a spectrum of technologies that reduce or eliminate manual handling across supply chain nodes. In warehouses, this includes autonomous mobile robots (AMRs), automated storage and retrieval systems (AS/RS), and robotic pick-and-place arms. In transportation, it extends to route optimisation algorithms, autonomous ground vehicles (AGVs), and drone delivery platforms. The common thread is the substitution of labour-intensive, energy-inefficient processes with digitally orchestrated, electrically powered alternatives.

Drone delivery refers to the use of unmanned aerial vehicles (UAVs) to transport goods directly to consumers or intermediate hubs. Modern delivery drones typically carry payloads of 2 to 5 kilograms over distances of 10 to 25 kilometres, operating on battery-electric propulsion. Their environmental advantage stems from drastically lower energy consumption per delivery compared with road vehicles: a delivery drone consumes approximately 0.08 kWh per kilometre versus 0.5 to 1.5 kWh per kilometre for an electric delivery van (Nature Energy, 2024).

Last-mile delivery describes the final leg of the supply chain, from the last distribution point to the end customer. This segment is disproportionately carbon-intensive because vehicles make frequent stops, travel short distances between deliveries, and often operate in congested urban environments where idling and stop-start driving patterns degrade fuel efficiency. Innovations targeting this segment include micro-fulfilment centres, cargo bikes, locker networks, crowd-sourced delivery platforms, and autonomous delivery robots.

Autonomous ground vehicles (AGVs) are electrically powered robots that navigate pavements or designated lanes to deliver packages without human drivers. Companies like Starship Technologies have completed over 6 million autonomous deliveries across 20 countries as of early 2026, demonstrating commercial viability at scale.

Micro-fulfilment centres (MFCs) are small, automated warehouses located within urban areas, typically 2,000 to 8,000 square feet, that bring inventory closer to consumers and reduce the distance and emissions of each delivery. By shortening the last mile to the "last block," MFCs can reduce delivery emissions by 50 to 75 percent compared with traditional suburban distribution centres (Deloitte, 2025).

How It Works

The modern automated logistics chain operates as an integrated system. Orders flow into warehouse management software that directs AMRs to retrieve items from high-density storage racks. Robotic arms pick and pack items into route-optimised containers. AI-powered routing engines then assign each parcel to the most efficient delivery mode based on weight, distance, urgency, weather conditions, and carbon constraints.

For deliveries within a 15-kilometre radius, drone dispatch becomes the lowest-emission option. The drone navigates via GPS and computer vision to the delivery coordinates, descends to a safe altitude, and lowers the package using a winch mechanism before returning to its charging hub. Wing, Alphabet's drone delivery subsidiary, has demonstrated this model at scale in the Dallas-Fort Worth metropolitan area, completing over 350,000 commercial deliveries by the end of 2025 (Wing, 2026).

For heavier or longer-distance urban deliveries, autonomous ground vehicles navigate pavements using lidar, cameras, and ultrasonic sensors. These robots operate at walking speed, drawing minimal energy and producing zero tailpipe emissions. In suburban and rural settings where neither drones nor sidewalk robots are practical, electric delivery vans equipped with dynamic route optimisation complete the network.

The entire system feeds data back into a central platform that continuously learns and improves. Machine learning models analyse delivery success rates, energy consumption patterns, traffic conditions, and weather forecasts to refine routing decisions, predict maintenance needs, and optimise fleet composition. Amazon has deployed this integrated approach across its delivery network, combining over 750,000 robotic units in warehouses with a growing fleet of electric delivery vans and drone trials under its Prime Air programme (Amazon, 2025).

What's Working

Drone delivery is achieving commercial scale. Wing reported a 300 percent increase in delivery volumes in 2025 compared with the previous year, completing its first 100,000 deliveries in Australia in just over two years and its second 100,000 in under six months (Wing, 2026). Zipline, which pioneered medical drone delivery in Rwanda and Ghana, has expanded into commercial delivery in the United States, operating in partnership with Walmart across seven states and delivering over 1 million packages in 2025 (Zipline, 2025). These deployments demonstrate that drone delivery can achieve unit economics competitive with traditional couriers in specific use cases, particularly for lightweight, time-sensitive goods.

Warehouse automation is delivering measurable sustainability gains. DHL Supply Chain has deployed over 8,000 robots across its facilities globally, reporting a 25 percent improvement in energy efficiency per parcel handled and a 30 percent reduction in error rates that cause returns and redeliveries (DHL, 2025). Ocado Group's automated fulfilment centres in the UK use robotic grids that process 65,000 items per hour while consuming 50 percent less energy per order than comparable manual operations (Ocado, 2025). The reduction in returns alone has significant carbon implications: the reverse logistics of returned goods generates an estimated 27 million tonnes of CO2 annually in the United States (Optoro, 2024).

Route optimisation software is cutting emissions at scale. UPS's ORION system, which uses advanced analytics to optimise delivery routes, saves the company approximately 100 million miles driven per year, equivalent to eliminating roughly 100,000 metric tonnes of CO2 emissions annually (UPS, 2025). Newer AI-driven platforms from companies like FarEye and Locus Robotics further improve upon these gains by incorporating real-time traffic, emissions pricing, and multimodal options into routing decisions.

Micro-fulfilment is reshaping urban logistics. Retailers including Kroger, Woolworths, and Carrefour have deployed MFCs powered by automation technology from Takeoff Technologies and Fabric, reducing average delivery distances by 60 percent and enabling sub-two-hour delivery windows without expanded vehicle fleets (Fabric, 2025).

What Isn't Working

Regulatory fragmentation is slowing drone adoption. Despite technological readiness, drone delivery faces a patchwork of airspace regulations that vary by country, state, and municipality. In the European Union, the U-space regulatory framework is still being implemented across member states, creating uncertainty for operators planning cross-border services. In the United States, the FAA's Part 135 certification process remains slow, with only a handful of operators holding full commercial licences as of early 2026 (FAA, 2025). India, despite launching a liberalised drone policy in 2021, has seen limited commercial delivery adoption due to local airspace restrictions and infrastructure gaps.

Payload and range limitations constrain drone economics. Current battery technology limits most delivery drones to payloads under 5 kilograms and round-trip ranges of 20 to 30 kilometres. This means that only 10 to 15 percent of e-commerce parcels by weight are drone-eligible, significantly limiting the addressable market (McKinsey, 2025). Heavier goods, multi-item orders, and deliveries beyond suburban boundaries still require ground-based solutions, reducing the potential emissions savings from drone programmes.

Labour displacement concerns create political headwinds. Warehouse automation and autonomous delivery threaten millions of logistics jobs globally. The International Transport Forum estimates that 3.5 million driving jobs in Europe alone could be affected by autonomous logistics by 2030 (ITF, 2024). Without proactive reskilling programmes and just-transition policies, political opposition can delay or block deployment of automated systems, as seen in California's legislative efforts to regulate autonomous trucks.

Infrastructure gaps in developing markets. Automated logistics requires reliable electricity, broadband connectivity, accurate digital mapping, and maintained road or pavement surfaces. Many markets in sub-Saharan Africa, South Asia, and Southeast Asia lack these prerequisites, limiting the applicability of warehouse automation, autonomous ground vehicles, and even drone delivery beyond narrow medical supply use cases.

Cybersecurity and data privacy risks are underappreciated. Connected logistics systems generate vast amounts of data about delivery patterns, consumer behaviour, and infrastructure. Without robust cybersecurity frameworks, these systems are vulnerable to attacks that could disrupt supply chains or compromise personal data. The 2024 cyberattack on a major European logistics provider that disrupted deliveries across 14 countries underscored this vulnerability (ENISA, 2025).

Key Players

Established Leaders

• Amazon — Operates over 750,000 warehouse robots and the Prime Air drone delivery programme across multiple US and international markets.
• DHL Supply Chain — Deployed 8,000+ logistics robots globally with integrated sustainability tracking.
• UPS — ORION route optimisation platform eliminates 100 million miles annually; piloting drone delivery with DroneUp.
• Wing (Alphabet) — Leading commercial drone delivery operator with 350,000+ deliveries across Australia, US, and Europe.

Emerging Startups

• Zipline — Autonomous aerial logistics across 10 countries; 1 million+ commercial deliveries in 2025.
• Starship Technologies — 6 million+ autonomous sidewalk robot deliveries in 20 countries.
• Fabric — Micro-fulfilment automation technology deployed by major grocery retailers.
• FarEye — AI-powered delivery management and route optimisation platform.

Key Investors & Funders

• Breakthrough Energy Ventures — Backing low-carbon logistics startups including autonomous and electric delivery.
• SoftBank Vision Fund — Major investor in autonomous delivery and warehouse robotics.
• Amazon Climate Pledge Fund — Investing in decarbonisation of logistics and supply chains.

Sector-Specific KPI Benchmarks

Action Checklist

• Baseline your last-mile emissions. Measure CO2e per parcel across all delivery modes to identify the highest-impact intervention points.
• Pilot drone or autonomous delivery in one geography. Start with a controlled market where regulations are favourable and population density supports viable unit economics.
• Invest in route optimisation software. Even without hardware changes, AI-driven routing can reduce fuel consumption and emissions by 10 to 25 percent within months.
• Evaluate micro-fulfilment for urban markets. Assess whether MFCs can reduce delivery distances and enable same-day service while cutting emissions.
• Engage with regulators proactively. Participate in airspace and autonomous vehicle consultations to shape rules that enable sustainable innovation.
• Plan for workforce transition. Develop reskilling programmes for warehouse and delivery workers who may be displaced by automation.
• Integrate sustainability metrics into logistics procurement. Require carriers and 3PL partners to report emissions intensity per delivery and demonstrate improvement trajectories.

FAQ

How much can drone delivery reduce carbon emissions compared with van delivery? Research published in Nature Energy (2024) found that drone delivery produces 30 to 80 percent fewer emissions per package compared with diesel van delivery, depending on the electricity grid mix and delivery density. In markets with high renewable electricity penetration, drone delivery can achieve near-zero operational emissions. However, these savings apply primarily to lightweight packages under 5 kilograms and short distances under 15 kilometres.

What are the biggest barriers to scaling logistics automation? The three primary barriers are regulatory fragmentation (especially for drones and autonomous vehicles), high upfront capital costs for warehouse automation, and workforce transition challenges. Regulatory uncertainty alone can add two to three years to deployment timelines, while warehouse automation systems typically require $5 to $15 million in initial investment per facility, with payback periods of three to five years.

Is autonomous last-mile delivery commercially viable today? In specific use cases, yes. Starship Technologies operates profitably in university campuses and suburban neighbourhoods, and Wing has demonstrated competitive unit economics for lightweight deliveries. However, broad commercial viability across all parcel types and geographies has not yet been achieved. Most operators still require subsidies, partnerships, or premium pricing to sustain operations while scaling.

How should companies prioritise among these technologies? Start with the highest-impact, lowest-risk interventions first. Route optimisation software offers immediate emissions reductions with minimal capital expenditure. Warehouse automation delivers the next tier of gains with moderate investment and proven ROI. Drone and autonomous delivery should be piloted in favourable regulatory environments and scaled based on demonstrated performance. The optimal mix depends on delivery density, parcel characteristics, regulatory environment, and available infrastructure.

Sources

• World Economic Forum. (2024). The Future of the Last-Mile Ecosystem: Transition Roadmaps for Public- and Private-Sector Players. WEF.
• Pitney Bowes. (2025). Parcel Shipping Index: Global Parcel Volume Projections 2024-2027. Pitney Bowes.
• McKinsey & Company. (2025). Autonomous Delivery and Logistics Automation: The Path to Scale. McKinsey.
• Nature Energy. (2024). Comparative Life-Cycle Assessment of Drone and Ground Vehicle Delivery. Nature Energy, 9(3), 215-228.
• Wing. (2026). 2025 Annual Impact Report: Commercial Drone Delivery at Scale. Wing Aviation LLC.
• Zipline. (2025). Annual Report: Instant Logistics Across Ten Countries. Zipline International.
• Amazon. (2025). Sustainability Report: Robotics, Electrification, and Last-Mile Innovation. Amazon.
• DHL. (2025). Global Logistics Automation and Sustainability Progress Report. DHL Supply Chain.
• UPS. (2025). ORION and Beyond: Sustainability Through Route Optimisation. United Parcel Service.
• Deloitte. (2025). Micro-Fulfilment and Urban Logistics: Emissions Reduction Through Proximity. Deloitte Insights.
• FAA. (2025). Unmanned Aircraft Systems: Regulatory Framework and Commercial Certification Status. Federal Aviation Administration.
• ITF. (2024). Transport Employment and Automation: Scenarios for 2030. International Transport Forum, OECD.
• ENISA. (2025). Threat Landscape for Transport and Logistics Sector. European Union Agency for Cybersecurity.