Not too long ago, Jeff Bezos sparked a tidal wave of controversy when he spoke of the potential of Amazon's 'Prime Air' drone delivery service. For those of you unfamiliar with the concept, for packages below 5 pounds ((about 2 Kg) around 86% of Amazon's
current deliveries) and with delivery destinations within 30 minutes - Amazon's drone army will be able to collect the package (using a cable/winch with a secure holding mechanism to maintain a safe height), draw the package up by withdrawing the cable, travel
to the delivery destination and lower the package and release.
In city environments, secure cages on top of certain buildings will be used for package deliveries, recipients will be able to retrieve their packages from these cages using secure codes (like Amazon 'smart lockers' today). For suburbs and rural environments
customers will be able to select a delivery zone where the drone will drop the package - e.g. your garden lawn/ local school sports ground.
This of course holds great potential, presenting us with a vision of motorways bereft of the diesel guzzling, CO2 spewing Lorries, vans and cars that are used to today to transport goods.
In addition the various safety concerns are quickly being overcome by talented developers who are using machine learning, image recognition, battery charge calculation, Geo - fencing and other technologies in order to ensure drones don't descend too far,
travel too fast, get lost, crash into anything/each other or run out of power. This is all in addition to the pilot's own intuition and skill.
But there is one emerging issue.
GPS is fantastically accurate when guiding a drone across a city, however it is extremely inaccurate when measuring altitude during a drones descent. In order to overcome this, drones are equipped with LiDAR, to sense height, and optics that recognise any
potential hazards that a pilot may encounter.
It is not the inaccuracy that is the issue, it is the use of optics. In order to ensure that the drones continue to improve on their safety intelligence after being released, Amazon will undoubtedly use machine learning to ensure that the 'drone network'
shares lessons, learns from mistakes and debug any sloppy code. This will mean that drones will most likely be connected to the cloud.
This, in turn, means that Amazon could potentially be coerced into working with Governments/ Police authorities to use this drone network as a new method of citizen surveillance(through composite images being combined on a large (city wide) scale). Monitoring
crowds, children's play areas, pub gardens and major shopping streets. It is predicted that, as with the rise of the automobile, though it may be a rarity to see a drone in the sky today, it will not be strange to see 30+ drones over your head in just a couple
of decades. This poses serious ethical questions about privacy that many are currently debating. There is a undountedly a need to develop regulation that addresses not just their safe flight, but also the privacy implications of having British skies filled
with hovering data-collecting robots.
Whilst in the short term there is potential for somebody to build a multi-sided market place for qualified delivery drone pilots and retailers (other than Amazon) - as drones become more autonomous and the human element is removed, this ethical issue will
only become more prevalent.
Is this a case of cost & convenience over ethics?
What do you think the future holds? Will the 3D printing revolution turn this problem on its head?