Amazon’s Drone Delivery Plans: What’s Old, What’s New And When?

Amazon’s Prime Air delivery drone has some great new technology but the service faces strong and persistent regulatory hurdles.

Amazon’s drone delivery service is in the news again, this time with a new drone named the “MK27.” As the latest iteration of Amazon’s Prime Air delivery drone, it’s apparently safer, more efficient, and more stable than previous models, according to Amazon’s CEO Worldwide Consumer, Jeff Wilke. Wilke also said the company expects to scale the Prime Air delivery drone quickly, with the hope that it will be able to bring packages to customers “within months.”

What’s old?

For some, this may sound like déjà vu, and they’d be right. In December 2013, Amazon CEO and Founder Jeff Bezos told “60 Minutes” that drones would be flying to customers’ homes within five years. But that deadline came and went due to the many regulatory and technical hurdles that drone delivery companies face. Since Amazon first announced its plan for a drone delivery service, the company has gone through more than two dozen drone designs, none of which was able to adequately avoid other aircraft, objects, or people on the ground.

The Federal Aviation Administration, which regulates commercial drone use in the U.S., has approved previous Amazon drones for test flights, but each new prototype needs a special airworthiness certificate. Notwithstanding, the FAA told Forbes it has approved for Amazon one year of research and testing, allowing the company to operate its new unmanned aircraft for research and development and crew training in authorized flight areas—though not for deliveries.

What’s new?

Still, to Amazon’s credit, they have been hard at work moving drone technology forward for all. This newest drone comes packed with some impressive features, including artificial intelligence that allows it to operate more autonomously. For example, if the drone’s flight environment changes while it’s in transit or it comes into contact with a moving object while approaching the delivery destination, it will either make an appropriate evasive move, delay delivery, or abort delivery altogether. For most drones, this kind of “sense and avoid” behavior is performed by a remote pilot. In Amazon’s case, the drone employs proprietary computer vision and machine learning algorithms that can detect people, animals, and even wires and know what to do about them as they descend into, and ascend out of, a customer’s yard.

The other “cool factor” of this drone is its hybrid design. As this video shows, it takes off and lands vertically, like a helicopter, and flies horizontally and aerodynamically like an airplane—and it transitions between these two modes. While hybrid aircraft are nothing new, Amazon’s drone avoids the common shortcoming of most hybrid “tail-sitter” drones like this one by including a vision positioning system to keep it stable in the wind when landing.

When can I enjoy drone delivery?

All said and done, you can scratch that “within months” phrase. Realistically, Amazon is looking at a couple of years for any regular delivery operations—and those would be restricted. Mind you, it will be restricted because the FAA has yet to work out solutions for old and persistent problems like identifying drones, flying them over people, flying them in urban areas, and managing unmanned aircraft traffic, to name a few. Until then, we’re likely to see more headlines. Who knows, perhaps you’ll be one the lucky first recipient to get an emergency package of diapers delivered to your back yard via Amazon Prime Air once they begin testing.

 

Image credit – Photographer: Joe Buglewicz/Bloomberg © 2019 BLOOMBERG FINANCE LP

This post first appeared on Forbes.com

Why the Drone Network of Tomorrow is Farther Away than You Think

Airspace integration and management solutions for drones continue to garner new investment, but most options are based on fairytale scenarios and raise more questions than answers.

I’ve been doing research on the commercial drone industry since early 2012, and it never ceases to amaze me how much hype there is.  A week doesn’t go by where I find a new fantasy forecast or see an announcement on how this or that drone networking solution is “game changing.”

How real are those claims that drones will one day be filling our skies and delivering packages? Where and when will we see massive industry growth and is that growth dependent on the existence of a drone network?  In this post, I’ll go over a few misconceptions, discuss the harsh reality, and offer two lessons learned that I hope will help make the conversation a bit more rational.

The hype

Question: How much spin is out there on drone networks?  Answer: A lot.

Take this piece, for example: In The Drone Network of Tomorrow (It’s Closer Than You Think). The author wants you to believe that the drone network of tomorrow is a few hurdles away.  In this futuristic world, users will remotely dispatch multiple drones right from their offices. They’ll specify the flight path, and the drones will fly there autonomously and collect data. In this world, there will be drones-for-hire stationed at key locations and you will just click on button to summon them at your command. It will be “the Internet of drones” and it will be accomplished via the LTE network, the same network to which every smartphone is connected today.

Investors buy it.

Read The Big Money Continues to Bet on Drones, which discusses Verizon’s recent acquisition of Skyward. Read Airmap’s own take on their announcement of $26 million in Series B funding from Microsoft, Airbus, Qualcomm, Yuneec, and Sony, with Microsoft leading the round.

The press buys it.

Read this recent article in Recode. It says:

Drones are, after all, flying computers that connect to the internet—connectivity on a drone is often used to share flight information with other drones, report to air traffic control or send aerial imaging back in real time to the operator.

I bought it, too.

In December 2014, I wrote Why Drones Are the Future of the Internet of Things.

But since that time I’ve done a lot a research to find evidence supporting industry claims, and the truth is, at every turn I’ve come up empty handed and found many misconceptions.

The misconceptions

Many in the industry have worked together to move forward the various Unmanned Traffic Management (UTM) projects. UTM refers to efforts to build an air traffic management infrastructure for drones, such as the NASA-FAA UTM project and GUTMA. Those initiatives are a collaboration between government regulators and private industry partners. At the center of those initiatives is the enablement of routine beyond visual line of sight (BVLOS–sometimes just BLOS) operations for commercial drones. Good. We need that. To see some of that work, download the latest presentations from the 2016 UTM Conference here.

While the NASA-FAA UTM initiative may have started out with some simple solutions, it’s now blossomed into an expensive “one-size-fits-all” behemoth that is proposing ways to control flight scenarios that don’t need them—those flights where no data exists showing any risk those operations pose to the NAS or nonparticipants on the ground. That hasn’t stopped UTM participants, along with the Drone Advisory Committee (DAC), though, from suggesting those controls.

I think it’s a good idea that’s gone bad. I am not alone in perceiving that many UTM and DAC participants think their charter is to integrate the Internet and the cellular network into the National Airspace System (NAS). The leader of GUTMA thinks his organization can do for drones what ICANN does for the Internet. Face palm.

Now, the snowball effect is these companies (and investors) believe drones are Internet-of-Things (IoT) devices that are going to magically multiply like rabbits once we have a drone network. Truth is, drones aren’t IoT devices; they’re data-gathering aircraft.  Yes, they collect data that looks a lot like the data from an IoT device in motion (see my presentation on that here), but equating them with IoT devices assumes way too much—like the need for constant connectivity to the Internet, for one. Here’s a clue: Drones from DJI, the dominant market share leader, don’t have it, nor do 99.9% of drones that operate in the NAS today.

The harsh reality

So, to put it bluntly, the vision of tens of millions of drones flying in the NAS alongside manned aircraft is overstated. Visionaries like to point to the 100,000 or so flights that happen today as the bellwether indicator of what’s to come. They point to the headlines that say package delivery drones will fill the skies as reality. But drone delivery has been seriously debunked, and the bellwether argument is a non-sequitur. The vast majority of the flights happening today happen in uncontrolled Class G airspace, happen around 200-300 feet above ground level, happen without any automated traffic control interaction, and happen without incident.

Also, our research says the growth of drone use by industries will be much more measured than the hyped growth figures that visionaries tout, and the bulk of operations will happen mostly as they do today within visual line of sight (LOS).  We don’t see huge volumes for BVLOS operations happening for many, many years—if at all. There are other factors hindering drone adoption beside regulations and air traffic control.  There are other reasons why companies will stick with incumbent technology like satellites and manned aircraft.  We have written much about hype in the drone industry, and if that’s new to you, then you can start your research with this SlideShare.

Two lessons to take to heart

Regulators and experienced military users know flying drones safely and securely in BVLOS operations is not easy. Because of its complexity, we now have a new term in aviation: Performance Based Navigation (PBN). PBN describes requirements for separating aircraft and avoiding collision. PBN is a combination of systems both on and off of the aircraft that affect its ability to navigate. And that takes us to our first lesson.

Lesson 1: Buried in the recent announcement of regular BVLOS flights of Aeryon SkyRangers at the Foremost UAS Range in Alberta was the fact that Ventus Geospatial had to meet very stringent criteria from Transport Canada (Canada’s civil aviation authority). The requirements prescribe a host PBN including:

  • Sense-and-avoid system to provide traffic separation and a means for collision avoidance. That system will have to detect the traffic in time to process the sensor information, determine if a conflict exists, and execute a maneuver according to the right-of-way rules.  That system must possess the capability to detect both cooperative aircraft (aircraft with a means of electronic conspicuity (transponder, TCAS, ADS-B, etc.)) and non-cooperative aircraft.
  • Some ground-based radar systems may be utilized to provide a means of meeting sense and avoid requirements.
  • Traffic Alert and Collision Avoidance System (TCAS) / Airborne Collision Avoidance System (ACAS) that employs a collision avoidance system with reactive logic, so that any maneuver resulting from a perceived threat from another aircraft will not reduce the effectiveness of a TCAS/ACAS resolution advisory maneuver from that other aircraft.
  • Automated Dependent Surveillance Broadcast System (ADS-B), with the caveat that ADS-B does not have the ability to detect non-cooperative aircraft, it is not an approved strategy, in and of itself, for mitigating the UAV sense and avoid requirements.
  • Use of multilateration, which is a type of secondary surveillance system that is based on the use of conventional transponders and stationary receivers that provide an aircraft’s position using triangulation principles.
  • Separation and Collision Avoidance Standard Operating Procedures (SOPs) addressing:
    • Take-off/launch and landing/recovery procedures;
    • En-route and terminal procedures;
    • Loss of control data link; and
    • Abort procedures following critical system failure.

I could go on, but you get the point. There’s no uber-integrated automation system for PBN and there won’t be for a long time.

Lesson 2: Major General James Poss writes about his experiences with the early days of managing military drone systems in It’s the Data Link, Stupid. He says:

A commercial drone ground relay LOS network would have advantages and disadvantages compared to the Air Force system. Advantages are that existing cell phone tower networks are ideally positioned to provide the backbone for ground relay LOS networks for commercial drones. Cell phone companies have already leased the land, dealt with all the FCC regulatory restrictions and, most importantly, obtained the spectrum that could be used for BLOS drone link operations. They also have world class cyber defense centers.  The disadvantage is that today’s providers struggle to get enough bandwidth for existing cell phone coverage as it is, their cell tower antennas point down to cover ground users, vice up to cover drones, and their 4G link reliability won’t be high enough to satisfy FAA BLOS requirements. For something as critical as BLOS drone control, the FAA won’t tolerate ‘dropped calls’.  It’s that reliability thing, again.

He goes on to say:

Problems would remain even with this type of system. Although low bandwidth directional drone links using commercially available spectrum wouldn’t use scarce 4G service spectrum, they would take up physical space on already overcrowded cellphone towers. Directional antennas are expensive and must be positioned carefully to avoid radio frequency interference.

And I’m afraid 5G (the next-generation of mobile networks beyond the 4G LTE mobile networks of today) isn’t a silver bullet. Right now, forecasters say 5G adoption will be slow and most of that extra bandwidth will be used by consumers for mobile video viewing.

But General Poss raises great points. Why would anyone invest in these systems if they’re riskier than either military drones or manned aircraft, particularly when the regulatory environment is unclear?  Just how would the FAA regulate a drone network? If the control portion of the network would be an aviation safety critical system, would the FAA even have the authority to regulate it when it’s the Federal Communication Commission’s charter to regulate cellular communications?

All I can tell you is, Buckle up and stay tuned in. It’s going to be a long bumpy ride.

Drone Delivery: By The Numbers

What do medicine, batteries, and forgotten anniversary gifts have in common?  They are the most likely items consumers will want to be delivered by drone once that service is available.  At least, that’s the verdict from the consumers I surveyed in August and September of this year. In my post Drone Delivery: How Much Would You Pay?, I ran a poll with three simple questions:

  1. What’s the maximum amount you would be willing to pay for a package delivered by drone?
  2. Which of ten items would you want delivered in 30 minutes?
  3. Under what circumstance would you need something so quickly that you’d pay top dollar for it?

These are the topline results.  You can see the companion SlideShare presentation with more complete graphs and charts of the data here.

The max you’d pay?  First, I wanted some base data on how much people would pay for drone delivery.  So, I asked if consumers were willing to pay for the service and whether they wanted to pay a flat fee or a percentage of the price of their purchased items.  More than three-quarters of respondents (82%) told us they would be willing to pay (vs. 18% who said they wouldn’t), and the largest majority of those who’d pay (62%) said they would prefer paying a percentage of the item’s purchase price (vs. 18% who said they would rather pay a flat fee).

When we asked those who were willing to pay for the service how much they would be willing to pay, we saw big differences in preference. For instance, as I mentioned, only 18% of respondents said they’d prefer a flat fee, and 80% of those people said they wouldn’t pay more than US $50 for delivery.  That’s not a lot more than express overnight delivery fees. I doubt these consumers will be using fixed charge drone delivery services.

For those respondents who indicated they’d pay a percentage of an item’s price, more than half (51%) said they would pay up to 10% of an item’s purchase price.  Most of the rest (43%) said up to 20%, and only 6% said up to 30%.  It seems that a percentage charge could leave a delivery service with losses if the delivered items aren’t high priced.

Items you’d want delivered?  Second, I was curious to know what items consumers would likely purchase and want delivered by drone. I reviewed online shopping trends to find most popular product categories and top purchase drivers – keeping in mind the items had to fit the following drone delivery requirements:

  • The order must be small enough to fit in the drone’s cargo box
  • The items must weigh less than 5 lbs.

There were clear winners and losers in the list of items consumers would want delivered in 30 minutes or less (see Figure 1).

FIGURE 1 – Which of 10 Items Would You Want Delivered in 30 Minutes?

Drone Deliver Fig 1

Respondents could choose as many items as applies.  Nearly three-quarters (72%) would use drone delivery to have a gift delivered.  Over half (57%) chose electronics.  Food and medicine were next highest on the list at 45% and 38%, respectively.  Batteries were the most mentioned item in the ‘other’ category.

Food may not be a good candidate for drone delivery.  Whether from the supermarket, mini-mart, fast food joint, or other restaurant, food is generally very inexpensive per pound.

Circumstances under which you’d pay top delivery dollar? – Third, I wanted to know under what circumstances consumers would need something so quickly that they’d pay top dollar to have it delivered quickly. I looked at the same online shopping trends noted above for the most popular categories and top purchase drivers. Results appear in Figure 2.

FIGURE 2 – Under what circumstance would you need something so quickly that you’d pay top dollar for it?

Drone Deliver Fig 2

You can see that respondents picked ‘emergencies’ at 66% and ‘in trouble / need it quick’ at 64% as the most likely circumstances that would trigger the use of a drone delivery service.  An ‘emergency’ is a subjective term that can mean just about anything, including the third most likely circumstance of a last minute anniversary or birthday gift – especially if you are desperate and have forgotten it.

Bottom line – From this small study I see three interesting insights:

  1. For the most part, consumers are willing to use a drone delivery service, but price matters.
  2. Consumers are not willing to pay more than 30% of the item’s value for that delivery – most no more than 20%.
  3. Electronics given as gifts will likely be the most ordered items.
  4. Consumers consider medicine and batteries important and urgent enough for drone delivery.

Here’s a simple example based on the data gathered: A consumer camera is ordered as a last minute gift and needs to be delivered in 30 minutes or less. The best enthusiast compact camera in 2014 is the Sony DSC-RX100M II Cyber-Shot Digital Still Camera. Its shipping weight is 3 pounds. At the date of this post, it lists on Amazon for about US $650. If the drone delivery service charges 20%, then the extra cost will be US $130.  If the drone delivery service charges 30%, then the extra cost would be US $195.  Keep in mind this is an example only.  We do not know how much Amazon Prime Air will charge for this or any other items – nor do we know what items actually qualify for the service.  Regular Amazon Prime users know already that not all items qualify for its two-day delivery service since they ship from contracted suppliers.  However, for comparative purposes today, if you were to order this item for U.S. overnight delivery to the West Coast, the charge would be approximately $34.

I believe drone delivery is best suited to high value, lightweight merchandise, including some items at Walgreens, CVS, or Rite-Aid, and a meaningful portion of the merchandise at many other large retailers. In addition to consumer products delivered to households, drones can provide fast and cost-effective deliveries to remote, hazardous, or hard-to reach areas.  Other applications for drone delivery include:

  • Relief shipments or medicines needed on-site quickly
  • Autonomous logistics within corporate structures or facilities (micro logistics)
  • Transportation network or logistics for last-mile delivery of critical components or raw material
  • Document delivery

As always, feel free to comment, ask questions or email us directly at info@droneanalyst.com.

Image credit: Shutterstock

Drone Delivery: How much would you pay?

I love the look of the Tesla Model S, but I’d never get one. Somehow, the opportunity to pay $100K to drive only 300 miles before spending 4 to 9 hours recharging does not appeal to me. I also love the idea of Amazon’s drone delivery initiative Prime Air.  But, like the Tesla, the opportunity to pay a lot of money for drone delivery of the latest camera just doesn’t make sense.

Amazon has successfully painted a picture of a future in which its drones push merchandise from its warehouses to nearby consumers in 30 minutes. But to qualify for Prime Air, the order must be less than five pounds, which, according to Bezos, includes 86% of the packages Amazon currently sells. The order must also be small enough to fit in the cargo box that the craft will carry, and the delivery location must be within a 10-mile radius of a participating Amazon order fulfillment center.

What is less than 5 lbs.?  Here’s some examples from some recent online shopping trends research:

  • Consumer electronics
  • Books
  • Clothing

Sure I get that instant gratification is clearly one element why people might use Prime Air and providing outstanding service is another; as is staying ahead of the curve with innovative delivery and order fulfillment. All highly significant points in their own right to meet Amazon’s goal: “to be Earth’s most customer-centric company.” But as this article says cost of transportation is another. In 2013 Amazon’s incurred overall losses of $3.5 billion related to shipping costs. How long can that go on?

The problem is that I think drone delivery is going to cost a lot – maybe not at first but eventually. If Amazon can charge $99 per year for Amazon Prime, which provides delivery within 48 hours of placing the order, how much can it (or others) charge for nearly instantaneous delivery via drone? How much would you pay?  Would $500 per year be out of line?  How about $1,000?  What about an extra $100 (30% markup) to get your new GoPro delivered before you leave for the airport in an hour?

I’m curious what you would pay. So I’m running a poll with three simple questions:

  1. What’s the maximum amount you would be willing to pay for a package delivered by drone?
  2. Which of ten items would you want delivered in 30 minutes?
  3. Under what circumstance would you need something so quickly that you’d pay top dollar for it?

To take the poll click this link: https://www.surveymonkey.com/s/drone-delivery

I’ll run this poll for a few weeks.  Check back here later and I’ll post the results.

As always, I’m interested in hearing from you.  If have something to say feel free to comment below or email me at colin@droneanalyst.com.

Image credit: Amazon