How Pairing Drones with AI, the IoT, and RFID Tracking Is Making Business More Efficient

September 04, 2017 by Chantelle Dubois

Mixing and matching tech domains can produce some amazing results and sophisticated solutions to today’s problems.

Drones plus new tech equals a new world of applications. Mixing and matching tech domains can produce some amazing results and sophisticated solutions to today’s problems.

Drones, in particular, have established themselves quite firmly as a tech domain with what appears to be endless possibilities. Here, we take a look at what combining drones with other popular tech domains can lead to.

Drones and Flexible RFIDs: MIT's RFly

Online retailers and postal services manage billions of packages annually. Currently, human employees must scan the RFID tag of each package using hand held device at close range. This is time-consuming, limiting, and prone to human error. Mistakes are regularly made, and items regularly misplaced and lost.

Recognizing this, researchers from MIT’s Media Lab explored the possibility of using drones to track and manage inventories of boxes. 

Drones used in workspaces with humans need to be lightweight and plastic so they're safe in case of accidents. So, rather than attach a potentially heavy RFID reader to such a lightweight drone, the researchers fitted an RFID relay transmitter onto the drone instead. Such a drone can fly around a warehouse, identify and scan RFID tags, and then transmit that information back to a reader for use by a human operator. Not only is this incredibly useful for tracking inventory quickly, but it can also help locate lost items in a warehouse otherwise inundated with boxes.

This advancement has the potential to reduce billions of dollars of lost inventory, a problem that plagues large retailers the world over. Combined with previously pioneered disposable printed electronics for packaging, inventory management might be primed for a change to operations that increases efficiency and cuts costs.

Drones and AI: Intel's Myriad X SoC

Intel is expanding its artificial intelligence portfolio with its Myriad X SoC (PDF), featuring a Neural Compute Engine capable of reliable operation with high or low power. The intent is to provide a capable chip for use in machine learning, computer vision, and autonomous applications. This specifically will have a useful utility with drone technology.


The Myriad X has a dedicated Neural Network accelerator. Image courtesy of Intel.


Online retailers, once again, intend to make use out of drones through drone delivery services. Instead of waiting overnight, or maybe up to seven business days, for a package to be delivered, drones can be deployed in mere hours.

To navigate autonomously, a drone’s onboard computer must be capable of fast and reliable decision making. The Myriad X can manage a maximum of four trillion operations per second.

Drones and IoT: Drones as a Service (Drones-aaS)

IoT has been a rising trend for the last several years and looks to continue its trajectory in 2017. Networking giant, Cisco, is looking at ways to combine drone technology to enhance IoT applications. Head of Cisco’s drone program, Biren Gandhi, presented at the Small Unmanned Systems Business Expo (perhaps better known as the sUSB Expo or the "Silicon Valley Drone Show") in May, explaining the benefit of this tech fusion.

One point he made was that drones are becoming a more important part of commercial technology, with Drones-aaS (Drones as a Service) applications being the next paradigm after SaaS (Software as a Service). New, mobile, autonomous data collection has many uses in security and disaster management, and will be especially useful when plugged into the Internet providing quick access to data independent of the operator's location.

As drones prove their usefulness across various fields, businesses continue to invest in making them faster, tougher, and more reliable. But the intersection of drones and other growing technologies is an exciting one. Drones have yet to meet their full potential, but pairing them with new innovations is already making businesses more efficient.


Feature image courtesy of DJI.

  • P
    patman October 23, 2017

    Love this article. So cool to see that neural networks are now being integrated in SoCs.

    Typo note: “SaaS” stands for “Software as a Service”, not “Software as an Application”. I imagine the same would apply to “Drones as a Service”

    Like. Reply
    • K
      Kate Smith October 23, 2017
      Glad you liked the article, @patman! Thanks so much for pointing out the acronym issue—I've changed it to be more clear. It's kind of odd but Biren Gandhi actually uses "Drones-aaS" for "Drones as a Service" in his presentation, so that's what I changed it to!
      Like. Reply