For years, one of the biggest challenges in delivering goods to consumers was the “last mile”—getting products from a company’s distribution center to the customer. Today, a related challenge is playing out in drone delivery: drones can get goods above a delivery site, but they still must get the goods to the customer safely, accurately, and reliably. Currently, there are four main approaches (exhibit), but none has captured the overall market because of various needs across global markets and the plethora of operators attempting to find the right fit for specific services and segments. Given the pros and cons of each delivery method—along with customers’ needs, evolving regulations, and technological innovation—drone OEMs, operators, and investors will need to think strategically and continue innovating to both serve their current target customers and position themselves for future growth.
The last-meter challenge in drone delivery is real. Players must deal with issues such as safety (both for the package and for people on the ground), security after it is delivered, noise, congestion, and the need to optimize network delivery nodes—all under regulations that continue to evolve. There is no single solution to address them all, and players are designing a range of technologies and approaches.
For example, Droneup, Flytrex, Manna, SkyDrop, Wing, Wingcopter, and Zipline’s latest generation drones use a tether. When the drone reaches its destination, it hovers at a low altitude and deploys a tether that lowers the package to the ground. This approach allows precise targeting, delivery location flexibility, reverse logistics potential, and ensures the drone remains safely above people and objects, though the tether equipment can add complexity compared to other methods. As a result, this method can excel in lower density urban (e.g., rowhouse rooftops and small gardens) and many suburban and rural environments.
Other companies—such as Antwork, Matternet, and Meituan—use proprietary ground infrastructure, such as a docking station to receive drone deliveries. This technology supports secure deliveries to specific multi-consumer locations (such as hospitals, office complexes, apartment buildings), though it is less convenient for customers, since they cannot receive packages directly at their individual homes. Although the fixed-node delivery network reduces flexibility, the secure docking stations enable delivery of items either of particularly high-value or that require chain of custody, such as medications or medical samples and tissues. Given docking stations’ advantages of serving multiple deliveries and consumers in close proximity and providing the potential for battery charging and swapping, drones requiring ground infrastructure are typically best suited for urban and suburban contexts.
A few companies, including Amazon Prime Air and Drone Delivery Canada, simply drop the package from varying heights. That can be less complex from a technology requirement standpoint (i.e., less parts and components for tether systems, no ground infrastructure), but it entails higher risks. There is potential for package damage and injury to people due to falling packages, requiring additional safety precautions. And there is the risk of the drone colliding with objects or people as it hovers, sometimes just within a foot or two, off the ground while dropping a package. This method is better suited for suburban and rural contexts where adequate delivery site clearances are available to ensure safe spacing between drone and the surrounding environment, and where delivery zones may be larger to allow for packages to be dropped onto less precise landing zones.
And some companies—such as Zipline, with an initial drone design that is still operating around the world—meet last-meter needs by releasing the package at altitude and slowing its descent via parachute, with ability to land a package within a space approximately the size of two parking spaces. This method enables aircraft designs optimized for longer range and reduced complexity, for example fixed-wing. At present, the parachute method is best suited for rural and some lower-density suburban environments where delivery zones are larger with lesser need for precision.
Considerations for OEMs, operators, and investors
As the segment evolves, multiple factors will need to be considered when selecting the last-meter delivery method. Some of these include:
- Interoperability and standardization. Companies that developed dedicated ground infrastructure could face a push to standardize key aspects of their equipment—such as landing docks or chargers—to accommodate drones from other providers, particularly in highly constrained and in-demand locations.
- Delivery range. Battery and powertrain technology currently limit drone delivery range. Certain delivery methods, such as landing, dropping packages, or deploying a tethered component, further limit drone ranges because of high power requirements during hover.
- Evolving regulations. The industry is certain to face regulatory changes and headwinds and, as such, OEMs and operators must be prepared to react. Municipalities could enact regulations regarding safety and noise. Strict noise regulations could severely restrict both landing and drop methods (potentially less so for tethering, depending on the hover height). Drone-deployed parachute deliveries are a great alternative from a noise perspective, but they are less precise and carry safety concerns.
- Customer and bystander experience. Customer experience is important in aspects such as package security and precision delivery. But the experience of bystanders may be even more important—especially in the near term, when drone deliveries will be highly novel and bystanders will significantly outnumber customers.
- Aligning growth aspirations with delivery capabilities. The match between delivery method and market will be crucial as each method has advantages and disadvantages. Some methods will be more suitable for broad expansion across geographies and applications, while others will be more niche. Furthermore, each method will have varying costs that will scale differently. For example, landing with fixed ground infrastructure will require new capital and operating expenses for each added delivery node, whereas other methods can more readily add new delivery sites with little or no added cost.
OEMs and operators will need to keep these factors, and others, in mind when choosing the right path for their unique context. For example, operators may opt for a fleet strategy, with different last-meter solutions to meet the needs of different customer segments and geographies. OEMs could design a single platform that is capable of multiple last-meter approaches. And some may focus on a single platform and method that is ideally suited for a specific—and valuable—segment.
The right method for a given situation will depend upon regulatory frameworks, customer needs and preferences, public acceptance of drone delivery, and a myriad of other factors. All four methods could end up scaling to meet the needs of various market segments across the globe. As drone delivery use cases and volumes continue to grow, only time will tell if drone companies converge to, or diverge from, a single last-meter solution.
Andrea Cornell is a consultant in McKinsey’s Denver office, Brian Miller is consultant in the Boston office, and Robin Riedel is a partner in the San Francisco office.