FMJ.CO.UK FOOD SERVICE CASE STUDY
FEBRUARY 2022 25
commercial catering.
Getting the project ož the ground says
Atalian Servest’s Chef Director Chris Ince,
required a huge team ež ort between
caterers and engineers to complete a ‘fit for
purpose’ ož er.
“Let’s face it we had some chefs and some
engineers where there hasn’t been a lot of
interaction before now so in the first few
weeks we were dancing around each other,
but it’s now very smooth and easy and the
key point is there is a lot of trust there.
“Our job was to ensure we reduced the
possibility of customer error, for instance
if you’re customising your own bowl
you don’t want a range of conflicting
ingredients, it all has to run on a theme
and be coherent, and so there was
an extensive testing process with
the team to ensure the ingredients
we wanted to put in would hang
together.
“For us it’s fantastically exciting
and a radically diž erent service
from what came before. What the
machine does is extraordinary and
where it is breaking new ground is that
the range of products and the ability to
customise your lunch is almost limitless. It’s
sparked new lines of thinking, new thought
processes, it’s been very stimulating, it’s
been motivating and we’re absolutely
delighted.”
Says Wragg: “Chris and his team have been
fantastic at working with us on what kind of
menu we should have, what works with the
type of proposition we have, and using what
Atalian Servest knows about staž eating
habits and how we can do it at a price that
works. Their expertise on everything from
compliance issues to health and safety took
us from the concept stage right through to
onsite, serving meals every day.”
With this kind of complex installation,
Wragg reveals in the early days a large
engineering team was deployed to ‘superservice’
the robot. The onsite team has been
reduced in recent months to one full time
support person and two interns, which he
admits: “is overkill but this is our prototype
and we want to support Atalian Servest.
In the long term a machine like this would
probably be working with one member
of staž back of house and portering and
one member of staž running the customer
interface.”
For his part Ince explains that from the
back of house perspective serving food via a
robot requires a diž erent kind of approach.
For instance, instead of humans providing
direct customer service they’re instead
loading ingredients into a robot, but he
reveals: “the process has been successfully
weaved into the production plan for the
kitchen, just with a few diž erent agreed
parameters and cooking methods.”
He also points out that because
the Semblr produces personalised
hot and cold meals, it delivers
complete accuracy of portion
size, and given concerns
about issues such as food
allergies etc, až ords total
traceability of ingredients.
“One of the benefits
is at the point of service
everything is segregated so
there is no danger of cross contamination
whereas in a traditional service there is
potential for error.”
DATA GATHERING
One of the most innovative features of the
robot is the use of AI to enable predictionbased
service, with the end goal being that
when there are hundreds of these machines
installed in diž erent sites you can aggregate
the data and map it against things like
weather forecasts, traž ic analysis, the day of
the month/week, where it is in the year, and
in doing so understand the food demand
patterns.
Explains Wragg: “In this way you can
forecast out what you’re likely to need
to make in the kitchen and order the
ingredients in advance that support that,
based on large volumes of data that you
are intelligently filtering in order to plan
forward.”
For customers the ordering process is via
similar QR based app system we’ve all got
used to in pubs and restaurants during the
pandemic. Throughout the Ocado ož ices
there are QR codes for staž to pre-order for
‘click and collect’ – enabling them to order
their meal and nominate the time they want
to collect it.
Says Wragg: “What that means is by
around 11 to 11.30 each day we pretty much
know the exact quantity of everything we’re
going to sell that day, so if we know we’ve
got a quiet day we can pass that back to the
catering team. We also get feedback from
customers in real-time on every order that
comes through and that’s already led us to
devise a new menu from the feedback that
“)or Xs itƉs IantasticaOO\
ecitinJ anG a raGicaOO\ Giff erent
serYice IroP ZKat caPe EeIore :Kat
tKe PacKine Goes is etraorGinar\ anG
ZKere it is EreaNinJ neZ JroXnG is tKat
tKe ranJe oI SroGXcts anG tKe aEiOit\
to cXstoPise \oXr OXncK is
aOPost OiPitOess”