plenty of repetitive manual tasks across many industries.
His examples were the Brompton bike factory in
Richmond, for example. Observes Chandra: “Where are
the robots? We thought everything was automated; that
was the perception, but when we started to visit factories,
we saw that they weren’t using robots.”
The move by firms such as Denmark’s Universal
Robots with its lower cost collaborative units has taken a
step towards making the lower volume tasks more viable,
Chandra admits, but adds that while such a robot might
be priced at £17k, there are additional costs (see
diagram, left) that push an installation to £50k+. Eva is
an out-of-the box solution that only requires simple fixing
and a gripper system. It is offered at a price, with an ease
of use and integration that makes robotic automation
attractive where even today’s lower cost cobot is overkill,
in terms of demanded performance. The central question
is how has it been possible to deliver such a robot at
such a price?
Well, it has taken the entrepreneurial pair four years to
get to where the company is today, but Chandra explains
that a robot is really essentially “a computer and a bunch
of actuators, or joints – a transmission mechanism, a
position sensor, a motor and a brake system”. So, the
company set out to deliver 80% of the performance of
current servos at 20% of the cost, using globally
available, standard, commoditised products to build a
new servo-system. Robots are a low-volume, specialised
industry with components made for it. So, for example,
Eva’s motors are those used by the drone industry –
a $1,100 drive has become a $280 assembly. The
company has three patents on its own gearbox design.
In association with that, assembly processes were
designed to build this servo at scale economically, with
Automata working with Tharsus, the company that built
Ocado’s warehouse robot, to manufacture Eva. Tharsus is
a subcontract design and manufacture business that
makes machinery for other people, with its manufacturing
operation in the Blyth, Northumberland. Automata
concentrates on design and development of actuation
systems, refining product performance and assembly
manufacturing processes through collected data/KPIs.
Looking forward, the company will work with partfeeding
system partners to deliver systems, channel
partners to deploy turnkey installations for customers,
machine tool builders to deliver automated machine
solutions and will develop more software in support of
robot-based automation. Basically, it will create a
hardware and software ecosystem.
Customer requirement will guide further product
development, which could go in a number of directions,
says El Sayed. But he stresses that the central tenets for
the company will remain to provide the most affordable,
easiest to deploy and program robot-based automation
hardware. It’s got off to a good start. ■
LEAD FEATURE THE PROMPT FOR ROBOTIC INNOVATION
Integration (design, fabrication,
installation, programming,
testing, training): £12k+
Robot: £17k+
PLC (programmable logic
control): £3k+
End-effector: £1.5k+
Part
Part presentation: £10k+
Safety system: £3.5k+
Mounting: £1k+
Total deployed cost:
£50k+
of folded aluminium to achieve that. “Basically, we studied
origami for a year,” Chandra explained at the July event,
adding: “We became obsessed with automation, which
allowed a four-person team to completely automate the
panel design process – take the shape, break it down into
panels, flatten them, define the sequence of build and spit
out all the data necessary to realise it in the real world.”
The idea after that was that robots would underpin
automation at the production end, holding and folding the
500 panels. “Turns out it took half an hour to fold a panel
by robot and five minutes by hand. We didn’t have labour
cost as part of the project, so that turned out to be us.
We ended up folding 85% of the panels by hand, manually,
and shipping them to Venice.”
CHALLENGE SPURS INNOVATION
A bitter start with robots, then, with the team thinking that
if robotic automation was the promised land, why had their
experience been otherwise? The thinking was that they
would go out and buy a robot, using their own money, and
see if they could improve on the initial experience. But
with a starting price of £20,000-25,000 and no defined
use case, it was a challenge, Chandra said. Why not build
one then? “We can construct buildings, how hard can
robots be?” he posed.
After four months, there was a 3D-printed prototype
using off-the-shelf electronics that could be transported in
a backpack and could be “set and running in five
minutes”. The company has used Autodesk Fusion cloudbased
software to support its efforts. Today, four years on,
it takes three minutes to get Eva running, but the product
is a bit heavier, so a suitcase rather than a backpack is
the transport medium.
Chandra and El Sayed are no crazy designers. They
have analysed the market and the place that Eva can
satisfy in the automation market. Number one, there are
The cobot is
just the start.
There are other
costs that must
be added to an
installation’s
total price,
Chandra
advises. Eva,
below, is an outof
the box
solution that
only requires
simple fixing
and a gripper
system
12 November 2019 www.machinery.co.uk @MachineryTweets
/www.machinery.co.uk