What if every delivery, installation and repair in the world
could be effectively automated, with no human presence required? Whether pizzas
or refrigerators, furniture or pharmaceuticals, smartphones or snail mail, what
if it could all arrive at your door and even be installed inside without any
visitors ever being present? What if your deliveries could find you on the
move, or if your purchase could be constructed close at hand in the first
place, even those requiring special skills to make? What would happen if some
repairs could be handled by unmanned repair shops or even some devices’ own
internal systems?
What would it mean if a rare piece of expensive technology
could be repaired immediately at a remote base or if someone in need of a
brilliant neurosurgeon or cardiologist could receive their aid while hundreds
of miles from the nearest hospital?
How would other fields change if critical personnel could be
rushed to the scene by the dozens or thousands without deploying one additional
person, bolstering emergency personnel handling disaster sites, forest fires, missing-person
searches, security threats or other crises where the available assistance is
often neither numerous nor expert nor remotely expendable? What if a host of
systems could be actively coordinated by just a handful of people to deal with
even more exotic threats from emergent pandemics to countering cyber-threats?
What if even more hostile natural environments, such as
space itself, could be opened up to development by the same form of automation
and oversight?
And what if the little human intervention involved
constantly trained the software to require less and less human assistance, or
even awareness, so that the entire system would continue to become even more
self-reliant and adaptable? And what if all of this was already possible using
off-the-shelf technology, available from numerous sources?
Because it is. The building blocks already exist and now so
does the invention that brings them all together… patent pending.
The details can be found below, in this introduction, and
the rest of the article expands upon them.
But for now, the question is not: What if?
But rather: What now?
Here follows a brief description of the means to automate
all of these technologies with no human presence and minimal human
intervention. Not every image sent to the Patent Office is included herein, but a few resources are included because they so clearly illustrate the emerging
technologies that can be incorporated into the system as described.
The system begins with its core purpose, automating
deliveries, installations, repairs, remote manufacturing and assorted services
for the commercial world, and naturally expands from there.
So how do we automate these basic elements of retail, and
then everything else? The key innovations required are self-driving vehicles,
autonomous robots, VR gear, haptic gloves, our existing 3D street maps,
smartphone video cameras and GPS technology. Other innovations can be
incorporated as they become available, but what we need we already have.
Right now, a delivery, whether it is a shipped package, a
store white-goods delivery or a basic installation, has certain elements in
common. Someone loads an item into a truck, someone drives that truck to a
destination, and someone unloads that item at the other end. Unfortunately, you
need at least one person on every vehicle, committed to being there for the
entire delivery route, which can easily last for several hours, especially for
multiple deliveries. If a second delivery person is needed, you have just doubled
your personnel commitment.
And yet most if not all of the actions required by the
delivery team can just as easily be handled by machines, and the few requiring
human oversight and awareness take up only a fraction of the time required of
these employees. The human element is still irreplaceable in many instances,
but whether their skills are very basic or incredibly refined, too many
employees have to be paid for little more than traveling to and from the task at
hand… because their time is valuable too. But what if bringing that human
element to the project took almost no time or personnel at all, by cutting out
the elements well within the programmed expertise of machines?
Let us begin by considering the basic commercial delivery.
A simple package delivery – whether a mailed parcel, a small
purchase sent from the store, or even a pizza – only requires that the item
reach a customer or their doorstep, and possibly for said customer to sign a
pad or pay for their goods upon arrival. Google’s automated vehicles obviously
have the potential to get goods very close to your doorstep. Google has made a
point of testing their cars with live, competent humans in the driver’s seat,
but inevitably, unmanned vehicles are in our future. A likely interim step for
the unmanned auto would be the remote-oversight compromise – having someone watching
remotely who is capable of stepping in and controlling the vehicle in real time
and absolutely capable of hitting a “Stop” button if it looks like a truck is
about to back over, say, a child, a pet or a petunia. Note, however, that
existing high-end vehicles are already introducing safety features that involve
the machine reacting automatically and faster than a human driver in order to
protect its occupants. People will soon begin to see human intervention as less
critical, except with regards to their own judgment calls, for example, what
parts of their yard and driveway are accessible and what parts are off limits.
Once a vehicle has arrived, the simplest delivery can be
handled by an ordinary automated cart with similar but simpler programming than
the code used by the primary vehicle. All the cart needs to do is load up its
package, drive out of the truck, and roll up to a door. Other actions, such as
asking for money to be deposited before the release of the goods, proffering a
digital pad for a signature, or knocking on a door, are trivial and can be
handled using existing technology. A recorded voice and digital display can
request payment, payment can be handled like a self-service checkout, and
raising and holding an attached pad or firmly tapping on a door with a padded
limb are very straightforward motions. The latter can use a rangefinder to help
gauge distance, but once programmers know the optimum speed, direction and
force for the motion, the cart should be able to perform it every time with
whatever limb is built in to handle it. If they were in some way challenged by
this motion, the same cart could always include a recorded knocking noise, a
voice calling out and/or a directional doorbell with the sound projected
towards the residence or business and muffled in other directions. Simply
putting the speaker deep inside the cart at one end of a cone whose wider open
end will face toward the door will create the directional effect with minimal
effort. Meanwhile, inside the delivery truck, packages would be pre-loaded into
racks that would transfer them onto the delivering cart.
Stairways pose a challenge for this automated-cart option,
which are addressed with more complex deliveries below. Before leaving this
relatively basic method behind, however, remember that walking down two or
three steps will be an acceptable effort for most customers. Also remember that
some smaller obstacles are not as impassable as we might imagine. With three
steps before our plucky little cart, simply dropping a short, extendable ramp
from the top of the cart and having a smaller cart sitting on top drive onto
the porch to knock on the door and deliver the item will do. Alternatively, the
cart might start a bit further back, lay down a temporary ramp, and roll up
itself. Either way, the cart will be programmed to accept some locations as
inaccessible, to label them as such in its corporate database, and to allow for
other delivery methods in those instances. Human oversight, in the meantime,
will allow for the same discrimination – if a cart can not recognize that a
location poses an insurmountable obstacle to its delivery, but is likewise
unable to deliver, it will signal its human overseers after a pre-set interval
(for example, a five-minute delay in a step in the delivery) and they will make
that distinction for it with the touch of button. Further, even before this
system becomes more fully automated, one human being can easily watch a bank of
monitors and hit a pause or abort button and even take direct remote control
over the cart as necessary. A more automated cart will be able to pause and
signal whenever it needs assistance or is merely “confused” (facing a challenge
not clearly covered in its programming), enabling human overseers to handle
even more simultaneous deliveries. If multiple overseers watch these
deliveries, focusing on those actively occurring, each employee can take
exclusive control of a specific delivery and pause others until they or another
overseer can intervene. By allowing multiple watchers to observe and step in as
needed, simultaneous issues become less of a problem, allowing a larger number
of otherwise autonomous deliveries to occur per employee handling oversight
duties.
The standard, automated delivery would include a prerecorded
message announcing the order’s arrival and a photo of whoever took the package
and possibly signed for it. Barring more difficult terrain or a delivery
requiring more than a simple hand off and a signed digital pad, one or two
powered carts could handle the simplest door-to-door deliveries.
But the real breakthrough in delivery options remains
humanoid robots. Research companies have already developed both quadruped and
biped robots capable of independently navigating rough, natural terrain. Their
independence is still limited to moving around autonomously and carrying loads,
but for our purposes, that is all we need. Once you have a human-shaped robot
capable of walking into a building while managing a load, you can let humans
handle the rest. Carrying loads, walking up stairs and regaining their footing
when pushed off balance are basic tasks which these robots have already
mastered. And Petman and Atlas, given an adequate power supply, such as modern,
compressed-natural-gas fuel cells, already fit the bill. They just need to get
a better look, a human perspective and a little guidance.
How? Here is where we integrate cameras and remote sites
most thoroughly into the program, allowing delivery personnel wearing VR gear
(such as the recent Oculus Rift equipment, or any of its imitators) and haptic
gloves to view what is going on and to direct the route of the robots. More
importantly, with a clear view of what is going on and haptic gloves, those
personnel can take their robots in hand and not only guide them, but take
direct control over their arm movements. By having the robot’s arms mirror
those of the human controller during this remote override, an operator watching
from the perspective of cameras built into or attached to the robot’s head can
control their every gesture. Hence, pulling out refrigerators and plugging them
back in becomes something easily handled by a person who does that all the
time.
The advantage to this is that your delivery personnel are
not spending hours upon hours each day either driving to delivery locations or
riding along with the driver. You only need their time and attention during
each stop. Once their part in a delivery has been concluded, they can go on to
overseeing the next delivery or waiting for the next chance to use their
skills. The truck is entirely capable of getting itself to its next location.
Once no human oversight is required to drive from one site to another, you will
never need to commit personnel for the entire length of a trip, but only for
the delivery of goods that takes place once your vehicle arrives. For all
deliveries or installations requiring this kind of brief, routine, but skilled
or semi-skilled work, you can cut your personnel requirements drastically.
The capital cost is always key, but remember the price of
information technology is constantly dropping, even as its power increases. A
company making deliveries already has trucks – automating them is a matter of
adding electronic controls and a copy of the software. By the time you are
ready to do this, even on a small scale, that technology will not only have
proved itself, its price tag will have dropped well below the cost of the
excess drivers, who probably don’t want to be hauling refrigerators
indefinitely anyway. And some of whom will still be doing so, just on the other
end of those cameras and haptic sensors, instead of in person, using their own
backs. And $15 an hour, for 40 hours a week, for 2 personnel represents, in 10
weeks, $12,000. Or in 50 weeks, $60,000. For 200 personnel working at that
pace, $6,000,000. Meanwhile, the cost of computer hardware continues to drop at
the pace of Moore’s Law, and electronics have experienced a similar and ongoing
decline in price for decades.
Phased in with the cart-and-van basic delivery and an aesthetically
pleasing look for your humanoid robots, you will have an extremely
unintimidating method of getting goods to your customers. The basic delivery
itself, however, may not be so trivial in terms of its impacts.
Wedded to an effective inventory count and a reliable
confirmation of ordered goods (which could similarly be enhanced by double
checking through video), the basic delivery will enable customers who want
something from the store but who do not have time to get it during their busy
day a means of ordering remotely while at one location, and then receiving it
at a place and time of their choosing, even if they do not know where or
exactly when they will choose to receive it later.
Handling all of these actions will be illustrated and described
in greater detail below as will the other services enabled by this invention.