Why package delivery drones are a long way off
Amazon announced with much fanfare in 2013 their intention to develop and launch a drone delivery service. Two years later, drones have bumped 3D printers out of the technological limelight. Drones already make scheduled deliveries: DHL employs them to deliver medication to a North Sea island. And now Amazon has released a new video showing their own drone transitioning from vertical to horizontal flight and successfully delivering a much-needed replacement soccer cleat, narrated by Jeremy Clarkson (who is known for his technical prowess). From all appearances drones are well on their way to populating the skies of our neighborhood, buzzing about delivering our online purchases. In reality however we are much further off than most would think.
There are a number of reasons the development of drones has taken off so rapidly in the past 5 years or so. Technology needed for drones has become cheaper and more accessible; microcontrollers such as the Raspberry Pi are now inexpensive but powerful. The explosion of the smartphone market has also allowed the mass production of miniature gyroscope devices, critical for drone stabilization and flight. Cameras have become lighter and higher quality, while lithium ion battery technology has become light and cheap enough to produce en masse. All these incremental technological improvements have allowed drone manufacturers to develop and produce consumer drones at a price within the reach of most individuals with some spare cash. The commercial application of drones is a natural progression. But are drones ready for unmanned high volume and large scale operations such as package delivery? Though possible in theory, there are a number of significant hurdles that need to be overcome in order for commercial drone delivery to be successful.
Weather is perhaps the biggest challenge facing commercial drone delivery. Commercial aviation has advanced to the point where aircraft are able to operate in most all weather conditions. Rain, sleet, snow, wind, cold; nearly all weather is safe for flying. But these capabilities are in part due to technology that does not translate to drones – for example, anti-icing capabilities. All commercial aircraft are equipped with ice protection systems of some type: either heating elements, bleed air, or mechanical ice boots which provide protection from ice build-up on critical flight surfaces.
Drones do not currently have the capabilities to deal with ice accretion as current drone-sized batteries cannot provide sufficient power. Imagine if there was a snowstorm before the holiday season. Online shoppers relying on last-minute drone delivery would be unable to receive their purchases via drone. Uptime is especially critical for these types of last minute deliveries, if a package fails to be delivered on time customers are much more likely to simply drive to the nearest store rather than risk never getting the item they ordered. Even if a company like Amazon was to offer an alternative form of delivery as a contingency, the additional costs involved in maintaining not only a drone delivery fleet, but a traditional fleet of delivery vehicles as well or contracting with another company for same-day delivery, may prove too expensive.
Another environmental hurdle is wind. At the slow speeds drones can currently manage, sustained winds of 15-20 mph would significantly reduce their range or prohibit them from flying altogether. Current drone designs are optimized for vertical flight; most are not designed for high speed and efficient horizontal movement. Even when configured for horizontal flight, aircraft have complex systems designed to cope with a crosswind component. These systems, though complex, still require a human pilot to intervene in case an issue arises, something an unmanned drone will not have. Imagine a drone attempting to land to deliver a package on a windy day, in a yard covered by trees and flanked by power lines.
Visibility is frequently an issue with aviation in general. Complex systems of radar have been developed to help guide planes as they navigate the sky and land. Instrument landing systems (ILS) allow planes to hone in on a transponder signal to calculate the appropriate glide slope for a safe landing. CAT III-equipped airports and aircraft even can autoland without any pilot intervention, with visibility as low as 150ft. Delivery drones would have to contend with much more challenging conditions. Residential homes don’t have an ILS or localizer for a drone to track. Even with GPS capabilities, a drone will have to guide itself to a safe landing spot once it arrives on site. Amazon’s drone appears to use picture-type beacon to help guide drones to a safe landing spot. But how many people will know how to properly place these locators, not to mention a situation in typical urban and suburban areas which have tree cover, streetlights and electrical poles? Add in other environmental factors such as inclement weather and low visibility and the difficulties for drone operation seem outsized for their benefit.
Cargo capacity is another obvious issue. Battery technology still limits the amount of cargo a drone can carry due to the power density of lithium ion technology. There is a reason aircraft don’t run on battery power: pound for pound, lithium ion power density comes nowhere near Jet A. This will limit packages to a few pounds or less. A quick survey of my own Amazon purchases in the past year shows only a few packages would qualify. Package dimensions greatly limit drone usage as well. Even if a package remains within the required weight, it still has to fit within the cargo hold of a drone (Amazon’s appears to be about the size of a shoe box).
As drones begin to populate the skies, precautions need to also be put in place to prevent them from interfering with each other and other air traffic. Although the video does briefly reference these capabilities, no details are given. TCAS (Traffic collision avoidance systems) which are required on commercial aircraft, require both radar and transponders to operate. It an incredibly complex, standardized system, which plots the course of surrounding aircraft and calculates if a mid-air collision is imminent. A simple GPS-enabled reference is not effective in these scenarios, nor is a cellular internet connection robust enough for drones to communicate with one another. Equivalent systems need to be tested, standardized, and certified for drones. Safety, especially when operating in low altitude in highly populated areas is critical.
Many comparisons have been made to commercial aviation, and despite technological advances, this past year 79% of flights arrived on time. Commercial aviation’s systems have taken decades to develop. Would drone delivery service succeed with only 79% of packages ordered delivered on time, not to mention the challenges above?
If all that were not enough, the continued controversy around the FAA and their approach thus far for drone regulation throws another wrench in the works. Policies around pilots needing licenses to operate drones, having drones always be within the operator’s sight, and other requirements seem draconian and uninformed. The simple fact is no one, including the FAA, has had to deal with drones until now. Technology has far surpassed the legislation needed to regulate their use (as is often the case), the FAA’s position is a cautious one which is expected when you’re dealing with aircraft regulation.
All this discussion is moot unless Amazon customers are willing to pay extra for drone delivery, and how much money Amazon is willing to spend to develop, implement, and maintain a fleet of delivery drones (not to mention the lobbying costs to change FAA regulations). It is still unclear how many online shoppers subscribe to Amazon Prime, estimates vary greatly however it’s speculated Amazon loses between $1 and $2 billion in shipping for Prime customers every year. Drone delivery will only increase this overhead. Even then the question remains: would a typical American family use drone delivery for the scenario laid out in Amazon’s most recent promotional video? I find this questionable at best. It would be much easier, less costly, and perhaps even quicker, to run down to the nearest shoe store and pick up another sports shoe rather than sit on your front lawn awkwardly waiting for a drone to perhaps arrive. Drone delivery may soon go the way of the consumer 3D printer and might serve Amazon as PR stunt rather than a new form of delivery.