Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:00):
(buoyant music)
- Hello and welcome to,"insight.tech Talk,"
where we explore the latest IoT, edge, AI
and network technologytrends and innovations.
As always, I'm your host,
(00:20):
Christina Cardoza, EditorialDirector of insight.tech,
and today we're joined by Autocanteen
to talk about AI poweredself-service solutions.
But before we get started,let's get to know our guest,
Sergei, from Autocanteen.
What can you tell us aboutyourself and the company?
- Hi, it's a pleasure to be here.
Thank you, Christina, for the invite.
(00:40):
Yeah, Autocanteen is a AIpowered self-checkout solution.
And so our solution mainlycaters for food service industry
and we use computer visionand machine learning
to identify meals or retail items
very, very quickly and efficiently.
(01:01):
As of today, we processmillions of transactions
yeah, for guests per month,
reducing, first of all the queuing,
because it's a very fastexperience and very fast checkout
and that empowers staff members.
(01:23):
We pretty much give theteam more eyes, if you like,
to do the job and yeah, processing power
and well, Autocanteen terminalsare fully localized and live
(01:43):
in the UK, and Ireland, and Canada, US,
Czech Republic, Lithuania, Estonia,
so that's our brief introduction!
- Great, I love how you guys are focused
on the food service industry
because when we talk aboutself service solutions,
(02:04):
sometimes we're thinking retail
to self checkout in healthcare,
self check-in, things like that,
but I haven't really seen it too much
in the food service industry
and this is not goingthrough the drive through
and having AI talk to youthrough the drive through,
this is something that ifyou're online buying food,
it's recognizing the food,
it's using all this technology
(02:25):
to help make the food service industry
run a lot more efficiently.
So that's where I want tostart the conversation today,
talking about the challenges
that are facing food service operators
in terms of the efficiency,labor cost, customer experience,
why self-service solutions
was a great opportunity for this space.
(02:47):
- Self-service solutions, theyhelp to obviously maximize
and enhance the capacityof teams during peak times.
And previously it wasn't before,
a AI self-servicesolution like Autocanteen,
(03:09):
it wasn't possible to do self-service
for transactions within hospitality,
because those products,
they're not labeled and thereare no barcodes on them.
But with the technologies
of like computer visionand machine learning,
it's possible to identifythem very, very quickly
(03:30):
in a very similar fashionthat human beings do
and present the total andyeah, to automate that process.
The food service operators,they face the challenges
because like I said,
that operationally the main challenge
is that people tend to get hungry
(03:52):
at a very similar time
and they all turn up for lunchtime
or for breakfast all at the same time
and yeah, it's hundredsor thousands of people
and obviously, the challenge here,
you need to serve themquickly with food, with meals
(04:14):
and then also to transact very quickly
so that the food doesn't get cold
and people don't waste time.
So that's the challenge, number one
and also, imagine if you'reshort on stuff that day
and then, yeah, it might beeven challenging to cater.
(04:36):
So we can talk aboutthe labor costs as well,
but yeah, it's always a present factor
for any operation, for any team,
but it's important to think
how you can utilize that capacity
in the best possible way
(04:56):
to enhance the speed and customer service.
So that's I think the main challenges
that the food serviceoperators, they tackle,
during, yeah, service times
and we, with our solution, we help them
to remove some of thatload on their shoulders
(05:21):
and help them to get through
that service quicker and more efficiently.
- So to set the stage,
because we're talking aboutsometimes in retail environments
or there's no barcodes
in some of these itemsthat people are picking up.
So I'm thinking it's lunchtime,
I run into a store to grab a sandwich,
I make myself a salador grab a cup of soup,
(05:43):
there's no barcodes or anything on that
and then I go to checkout
and pay for the meal that I just created
and having to tell the foodservice operator at the end,
what is it that I'm eating,how much of it have I got
and that will bog up theline for other people then,
it doesn't make it aquick grab lunch and go
type of situation, is that correct?
(06:05):
- Absolutely, that's exactlythe process that you described.
And actually, like the traditionalmanned checkout points,
we know these metricsbecause we measure them
and traditional checkout points,
they take up to 30 seconds per transaction
(06:28):
and that is not very fast,
so that means you gettwo customers per minute
checking out via one checkout point
and we can do it with automatedsolution times faster,
so it's three to four times faster,
we can see that one checkout point
(06:50):
will process four to sixcustomers per minute.
And that's where the automationreally, really shines
and that's where Autocanteen solution
really can bring the benefit
because otherwise it's tricky
to get the same speed and efficiency.
(07:11):
If we talk about the times
and if we talk about the performance,
so our terminals, if welook at the transaction,
they will identifyeverything within a second.
So as you place the tray or retail items,
the algorithm will take the inputs
(07:32):
and within a second it'll promptyou the total on the screen
and then there is a couple of seconds
to acknowledge the totaland then pay for it,
so a few seconds to processtransaction and so on
and we end up with asub-10-seconds transaction
(07:53):
from start to finish and with the receipt,
so it's the components, yeah,
within that transaction time.
And some of the sites that we support,
they process 2,000 people,or sorry, not, "process,"
they cater for 2,000 people
(08:16):
and imagine what kind of operation it is
to feed everyone withinone hour or two hours
for that sort of capacity
and on those sites wecan see 20 to 30 people
(08:37):
checking out per minute with our terminal.
So we can see that 20to 30 transactions land
within the minute at peak time
and to achieve the same speed of service,
it would require a lotmore manned till points,
(08:57):
so yeah, the numbersjust talk for themselves.
And also another point of viewhere and additional benefit
is that the customerexperience in this journey
is if you can get throughthe checkout process
so much quicker, it onlyincreases your satisfaction
(09:21):
because you do it in a veryquick and efficient way
and your food doesn't get cold
and yeah, it's a nice magic
as some people see it, magic machine,
that processes your transactions
and the friction in this way,
(09:42):
so that's definitely a benefit
to the customer service in there.
From staff point of view,
imagine how powerful they can feel
when previously they could only process
two people per minute,
but now they're just keepingan eye on the process
(10:04):
that actually processes 20 to 30 people
and they can do it all by themselves,
just keeping an eye on the flow.
So there are multiple benefitsof such implementations.
- Absolutely, I can imagine
if I wanted to grab something to eat
and it took a long linefor me to get my food,
(10:25):
next time I grab something to eat,
especially when you're comingfor lunch on a rush hour
and you have limited time,
I'm probably going to go somewhere else
if it takes too long at one place.
So I can see the benefits,
both from the customer side andboth from the business side,
the customer getting theirfood in a timely manner,
in a manner that it'snot cold by the end of it
and then the business beingable to retain those customers
(10:47):
for their next purchases.
So that definitely soundslike a great solution
in the food service industry,
I'm wondering if you can tell us
about some of the technologythat's implemented
in the Autocanteen solution
that makes this automation possible.
- Yes, absolutely, so thankyou for that question.
(11:10):
We rely fundamentally on,yeah, like I mentioned,
computer vision and machine learning
and so computer vision feedsthe input into the algorithms
and then we do the analysis,
the decomposition, and then classification
(11:31):
and learning and so on.
And the terminals, theyare all interconnected,
I mean, within the samesite, within the same account
and once you've taught one machine,
your other machines obviouslyfollow the same knowledge
and all the terminals,
(11:54):
they're managed via aweb-based admin panel,
so you can make anychanges to add products,
pricing amendments or view reporting,
see how your sales going,
so that's all available at your fingertips
(12:15):
and yeah, you get theinformation synchronized
pretty much instantly.
So yeah, that's how thetechnological offering looks like
and yeah, those are fundamental components
(12:35):
of the offering.
- How do you work with technology partners
in the ecosystem to make this happen?
And I should mention
"inside.tech Talk," andinside.tech as a whole,
we're sponsored by Intel,but I'm curious how,
because I know Intelhas done a lot of work
when it comes to being able to bring AI
(12:57):
into these different solutions
and their hardware is making it possible
to do this all in real time,
make sure that it's highperformant, high efficiency.
So what is the valueof leveraging partners
and technology from partners like Intel?
- Yeah, we are super thankful to Intel
and to this partnership
because we have been relying
(13:20):
on their components withinour software and hardware.
And one of the componentsthat we use is OpenVINO™,
it's blazing fast and we canpretty much make decisions
and those computations on the fly
and it really shows the difference,
(13:44):
comparing it to other, thistoolkit, OpenVINO toolkit,
to other frameworks that we tried before
and yeah, we can only recommend it highly.
So yeah, OpenVINO hasbeen an integral part
of our solution, for sure.
And apart from thispartnership, we have others
(14:13):
that help us with payment processing
and on the hardware side work
with vendors such as Elo Touch
they have also great products
and great platform that we run on,
which also has Intel processors
(14:33):
and yeah, we prefer torun our applications
on those processors.
So yeah, this Intel partnershiphas been super helpful
during our journey.
- Yeah, that's great to hear.
(14:54):
One thing that I love
about the partners that you just mentioned
is that they're always lookingat what's on the forefront,
so they're always stayingon top of the latest trends.
This AI technology spaceis changing so fast
that they're always making sure
that they're providingtechnologies and capabilities
that allow their customersor their partners
to future proof and toscale and be flexible
(15:17):
to any innovations that arehappening in the industry.
I'm wondering if you can shareany customer examples with us
that highlight how Autocanteen works
in the food service industry,
you don't have to nameany names if you can't,
but just to give us a picture
of the real benefits and the value
(15:38):
and the impact that it's brought
to industries and to customers.
- Yeah, absolutely, I canshare some case studies.
So one site, for instance,in central London,
it processes about a million pounds a year
and serving hundreds ofthousands of customers
(16:01):
and that site alone showed the dynamics,
like I said before,
that it processes 20 to30 customers per minute
and also it saved a lot of labor efforts
needed for the transactional function
(16:26):
so they can focus onbetter customer service
or be really helping elsewhere
apart from being on the tillsand doing their routine job.
And also we can confidently say,
(16:48):
it was measured that the queuing,
so the solution saved thousandsof hours in queuing time,
so yeah, that's a greatKPI that we also rely on.
And in terms of case studies and names,
(17:08):
actually I would be proud
to share the name andalso the award they won.
So Restaurant Associatespart of Compass Group
and they are a great partner in the UK
and they have just wonawards at the Cateys,
prestigious award ceremony in the UK
(17:31):
within the food service industry.
And the award was,"Best Use of Technology"
by a food service operator,
and that was for theimplementation of Autocanteen
and of our solution.
So that was fantastic recognition,
obviously for within the industry
(17:53):
and it was the best solution
on, yeah on the markets.
- Yeah, all the benefits thatwe've been talking about,
all the improvements,it's no surprise to me
that they would win an award
for best use of innovative technology
(18:15):
because that's somethingthat's really solving a problem
in their industry, solving pain points.
We see so many times companiestry to implement technology
for the sake of implementing technology
and this is more ofimplementing technology
for the sake of easing pain points
and improving customer experiences,
so that's great to hear.
Congratulations to Restaurant Associates
(18:37):
and Autocanteen as wellfor making that happen.
- Thank you, thank you, Christina.
Yeah, that was great, greatnews for all of us, for sure
and yeah, we just welcomethose awards to keep coming.
- So how do you envision then,
this food service industry changing?
(19:00):
How will you continue to stay up
on any of the latest trends
and bring more innovationsto the forefront
that these awards and these innovations
and these recognitions justcontinue on your journey?
- Well, the food service industry,
I mean, if we look at itthrough the prism of automation
(19:24):
and what's going tohappen in the next years
and how these things aregoing to be optimized,
of course we cannot say
that every function isgoing to be automated.
because hospitality,
I mean, it cannot run withoutpeople and it shouldn't,
I'm a big believer about it.
(19:45):
However, those functions
that are very repetitiveand are monotonous,
where they can be automated, they will,
and we are bringing this
for the function of transactional function
and we're bringing automation
and we can also see these efforts already
(20:08):
in the kitchen, back of house.
So for example, for analysisfor waste management
and there is actuallysome robotic companies
that are challenging,not only challenging,
but already offering thesolutions how to make meals
and prepare pizzas and so on.
(20:32):
So the repetitive, yeah, tasks,
they will be automated, for sure,
but yeah, to a certain extent,
technology is not going toreplace people in hospitality,
that's for sure.
- Yeah, I think that's areally powerful statement too,
because hospitality and these experiences,
we still need that human elementand that human interaction,
(20:53):
but this is just enhancingthose experiences
to make the outcome andthe solution better,
so it's all exciting things.
I know you were talking about
how this technology is being used
in the back of the house too
and some robotic companyis trying to figure out
how they can use technology to make meals,
but I'm curious, beyond food service,
(21:13):
are there any other industries
that Autocanteen'sself-service capabilities
is going to be appliedto or any other areas
that we can expect tosee Autocanteen automate
some of these tasks?
- Absolutely, absolutely,
we are already helping on some sites
within retail and micro markets
(21:33):
to also enable, first of all,
either fast transactions orjust enable 24/7 capability
for yeah, unattended transactions.
So for example, micro markets
where you just havesome products around you
(21:54):
and there is a couple of terminals
that people or guests,customers can always use
and those are the environmentsthat we also help,
that we bring the benefitswith our solution.
(22:17):
Our journey, it started in 2020
and it was during pandemic year
and yeah, we started working with Aramark
and they, at the time, they were looking
for a fast and touchlessAI self checkout solution
and we had just that
and fast forward fouryears and our terminals,
(22:39):
they're helping operators
such as Compass Group,and Dussmann, and Delirest
and others to enhance their food service.
Within banks, insurancecompanies, Ministry of Defense,
factories, warehouses,entertainment parks, to handle peaks
(23:01):
and yeah, I think that's 24/7 capability
and impress customers.
So it's been a great journey.
- Great, well I can't wait to see
how Autocanteen continues totransform other industries
as well as see some ofthis technology come,
(23:21):
see it out in the wild
and come to my ownstores and retail places.
So, exciting to see youguys continue your journey.
I recommend our listenersvisit Autocanteen website,
get in touch to see how they can help you
solve some of the painpoints that you're having.
So this has been a great conversation.
Before we go,
(23:42):
I just want to throw itback to you one last time,
if there's any finalthoughts or key takeaways
you want to leave ourlisteners with today?
- Oh, I just wanted to say thanks to Intel
for the support on the journey
and for being such an amazing partner
and yeah, to contributeto our solution in a way
and yeah, so thank you.
(24:04):
- Great, well thank you for joining us
and for giving us someinsight into Autocanteen
and how you guys are serving up the future
of the food industry.
So appreciate your time.(buoyant music)
Thanks to our listenersfor tuning in with us.
Until next time, this hasbeen, "insight.tech Talk."
(24:24):
(buoyant music continues)