amazon connectivity drones

Connectivity issues and the concerns no one is talking about around drone delivery

The following piece is a guest post by telecommunications specialist George Smith.

By now it’s pretty clear that Amazon’s master plan to take their delivery services to the skies via drone isn’t going to be easy. There are issues around flying beyond line of site, implementing obstacle sensors to avoid collisions, and getting regulatory approval. But there’s a completely different issue that no one is talking about: connectivity.

Maintaining a strong and stable internet connection for drones

Since delivery drones would fly autonomously rather than have a designated pilot, each drone needs to have the ability to send and receive information to air traffic control instantaneously so they know which parts of the air to avoid. To do that, Amazon’s drones will likely utilize a mixture of Wi-Fi and cellular connectivity.

Cellular connectivity will likely be delivered using roaming M2M SIM cards. M2M (machine to machine) means the communication between two or more devices without need for human interaction; in most cases this communication is in the form of data exchanges over a cellular network. These SIM cards allow technologies like drones to monitor networks for the best connection wherever they are in the country.

Battery life of drones

But with M2M SIM cards comes another obstacle: battery power. Although M2M can provide drones with connectivity, they can also be power-thirsty if exchanging large amounts of data. A potential solution? Low-powered wide-area networks, also known as LPWAN. LPWAN is a type of wireless telecommunication network designed to allow long range communications at a low bit rate, meaning they are extremely power efficient.

Although LPWAN services such as SigFox and LoRa are becoming more widely available, they aren’t yet being implemented in technology that flies — but that doesn’t mean they never will.

Managing drone GPS connections

As well as a strong internet connection, delivery drones require GPS signals to pinpoint their location and allow the companies to monitor the locations of their drones. But as drones fly farther away, with ‘beyond the line of sight’ comes greater risk of a dropped GPS signal. GPS signals are typically lost because of bad weather, changes in landscape and tall buildings in urban areas.  GPS uses three or four satellites for extended periods of time at 50 bits per second. If Amazon want full visibility of their drones, they need to come up with a solution that helps avoid losing GPS connectivity whilst maintaining low battery consumption; a task that can be extremely difficult to implement.

One way to overcome this issue could be to prioritize the technologies being used at any one time. This could allow the GPS to work in bursts instead of on a continuous basis to show the general path of the drone, whilst other technologies such as obstacle detection are still in use. If the drone was to stay static — or go wildly off-route — for a few bursts, Amazon would be able to recognize there is an issue and retrieve the delivery drone whilst maintaining GPS connectivity and lower power consumption.

D2D communication

Drone-to-drone (D2D) communication is essentially M2M (machine to machine) communication between two or more drones, giving them the ability to talk to one another. This enables them to avoid other drones and therefore collisions.  This process requires frequent data exchanges so that every drone understands the airspace it can use.

Related read: How to start a successful drone aerial photography business

Obstacle detection

Drones need to monitor their surroundings in order to avoid any objects or obstacles that are in their flight path and aren’t transmitting any wireless signals. Though, many consumer drones (including the DJI Phantom 4) already have obstacle detection in their repertoire of functionalities.

If Amazon wants this technology in their drones, they need to be issued with sensors and software that can identify foreign objects and find alternative routes without disrupting their fellow flying delivery robots, hence Amazon’s keen interest in utilising “non-collaborative, sensor-based sense-and-avoid systems” and D2D within their delivery drones.

This is a guest post by George Smith. Smith is a Digital Marketing & PR Assistant for telecommunications specialist DuoCall.


  • Drone Minds says:

    Well, obviously Amazon drone deliveries aren’t planned anytime soon, not wide scale anyway. The initial idea and patent application was more of a publicity stunt I think, but now it has sort of gained momentum, especially as other companies and jumped on the bandwagon.
    It’s difficult to speculate on future technology when all we have as a reference is current technology. Drone tech is moving very fast, we know that, but we can’t know the full impact of future tech, either known like 5G networks, or unknown.
    Future drones will need to be more self-aware, and NOT rely so much on external input from 3rd party computing like GPS and phone signals, things that can be interrupted or dropped altogether.
    As of now, you don’t absolutely need a connection 100% of the time to plan and implement routes – they can be programed in at the outset, using internet connections, and remembered for the duration of the journey without needing to be connected the whole time.

  • Pete says:

    I believe this article is partly inaccurate and partly just incomplete. So I’ve written up a blog post about this, to clarify some points stated in this article here.

    Click my name here to go my blog post.

    To sum it up: Battery usage is just outright wrong for 4G/LTE connections. GPS coverage is generally quite well and alternatives are available. All in all, it is obvious that the article above was written by someone who does not have knowledge of this domain. There are a lot of marketing keywords in it, which probably improve SEO ranking. However, the intrinsic content of the article could have been much better. No offense 🙂

  • Eric says:

    The amount of power required for communicating between machines is insignificant compared to the power required to keep a drone in the air. I would highly doubt that it would impact your flight time at all.

    • Drone Minds says:

      This is true at the moment. I guess if they were a lot lighter, the power needed wouldn’t be as much. However, battery technology is also advancing very rapidly. My current drones fly for almost 30 minutes, and do a lot more things than the drones I used 6 years ago, which only flew for 6 to 8 minutes. There are many exciting advances in battery technology being developed right now, and who knows where thy will be in another 6 years?

      • Eric says:

        It will always be true, even if drones get lighter. The amount of power required by a communications circuit board is probably in the milliWatt range … maybe up to a couple Watts at the most. A Phantom 4 has a battery capacity of 81.3 Watt-hours, which if you assume 28 minutes flight time that comes out to 174 Watts of power consumed on average over the entire flight. So we’re talking a fraction of a percentage of the power would be used by the communications. Even if drones were able to be half the weight, the communications is still going to be so insignificant that it will be negligible.

        • Drone Minds says:

          I know, I agree with you. It’s not a factor that really needs much consideration. Even less so seeing as batteries are only getting better?

Leave a Reply