First published in AMT Magazine June/July 2017 issue

Remotely piloted aircraft (aerial drones) were traditionally limited to military applications and the like but are increasingly being developed for commercial applications including point-to-point delivery of goods.

Amazon has been at the forefront of experimentation and development of delivery services using drones. This is reflected in the number of drone related patent applications filed in the United States by Amazon. Innovative concepts described in Amazon patent filings include:

  • using a blimp as an airborne distribution centre from which drones are deployed to transport goods to users (US 9,305,280);
  • drone landing platforms on top of buildings, street light poles, phone towers, etc to enable drones to recharge/refuel, collect and drop off packages (US 9,387,928), and
  • delivery drones which can in-flight land on consenting transport vehicles such as trucks and buses and recharge using the airflow generated by the vehicle in motion (US 2016/0257401).

X, a Google affiliated company, has also been actively working on drone development. In recently granted US 9,580,173 (pictured right), Google described a payload delivery system suspended from a drone.

The system includes thrusters to compensate for cross winds during descent from the drone. Other companies actively seeking to patent their drone-based delivery innovations in the United States include IBM, Walmart and Qualcomm.

More locally, Australia Post conducted closed-field drone trials in April last year in a collaboration with ARI Labs, a Melbourne robotics company. The trials sought to investigate the feasibility and safety of drones for small parcel delivery. Australia Post reported late last year they were seeking permission from the Civil Aviation Safety Authority (CASA), the governing body of drone operation in Australia, to conduct further trials to test improvements in technology.

In this respect, special approval is required to operate outside of the regulations, which notably requires operators to at all times maintain sight of the drone during flight, and operate only one drone at a time. The rules governing drones in Australia were most recently amended in September 2016 with the removal of the need for an operator's certificate or a remote pilot licence for commercial use of drones under 2kg. However, the requirement for drone pilots to maintain line of sight remains a major obstacle to drone based commercial delivery services.

The safety of drones is a key concern and proven technology solutions will be needed before government bodies can even contemplate approving beyond light of sight drone flight, particularly over densely populated areas.

In a Senate inquiry early this year on the regulatory requirements of drones, Telstra submitted that its mobile network infrastructure could be harnessed to allow drones equipped with sensors to reliably communicate on-board sensory data and receive control inputs in real time, potentially paving the way for drones to be safely flown beyond visual range. Telstra is reportedly now working with CASA to demonstrate the ability to safely operate drones beyond visual range utilising its mobile network and sensors.

Although drone technology will no doubt improve with time, parcel size and weight contraints will likely always limit drone delivery to small, relatively lightweight packages. For larger and heavier items, automated delivery cars and trucks, which are able to sense their environment, avoid collisions, make decisions and navigate without any human input, may eventually become a common sight on our roads.

California is a focal point for the development of self-driving technology with traditional car companies competeting with technology giants such as Google, Apple and Uber, and artificial intelligence startups like to dominate the market. Founded in 2015, is taking a slightly differnent approach to other companies by focusing on creating artifical intelligence software for autonomous vehicles using deep learning to create a system that can learn behaviour and recognise patterns and then make appropriate decisions to driving scenarios not encountered previously.

It may not be well known, but IBM is also actively involved in autonomous vehicle development and has recently been granted two patents in the United States, namely US 9,361,409 and US 9,566,986, in which cognitive computing strategies are employed to model human behaviour and help autonomous vehicles to better anticipate the actions of other road vehicles driven by humans. IBM was granted over 8000 US patents in 2016 and has top the list of patent recipents for the past 24 years.

With so many companies investing big dollars in the development of drones and autonomous vehicles, it will be fascinating to observe the impact this disruptive technology will eventually undoubtly have over the next 10 to 20 years in transport and logistics.