🚗 WEJO & PLTR Stocks: Betting BIG on Connected Mobility

00:00
the world is changing our devices are
00:02
becoming more connected and data is
00:04
driving nearly every aspect of our daily
00:06
lives from our time at home to our time
00:08
at work and everything in between that
00:11
in between is where a company called
00:13
wejo comes in wijo is a newly publicly
00:15
traded company focused on connected
00:17
vehicle data wijo works with palantir to
00:20
organize trillions of data points from
00:22
over 11 million vehicles on the road
00:24
today then they use that data to help
00:26
their clients make strategic and
00:28
tactical business decisions imagine
00:30
being able to look at the driving and
00:32
charging patterns of electric vehicles
00:34
to figure out the best places to build
00:36
new charging infrastructure imagine
00:38
seeing the flow of traffic in real time
00:40
and managing traffic lights to shorten
00:42
everyone's commute based on what's
00:43
actually happening on the road imagine
00:45
using every car's temperature sensors
00:47
fog lights windshield wipers and even
00:49
their braking patterns to identify bad
00:51
weather and individual hazards on the
00:53
road well you don't have to imagine
00:55
these things because they're exactly the
00:57
kinds of capabilities that we joe is
00:59
building on top of palantir's foundry
01:01
richard barlow is wijo's founder and ceo
01:04
the last time we spoke was about a
01:06
quarter ago right before their business
01:08
combination with virtuoso acquisition
01:10
corp since then the market for high-tech
01:12
growth stocks has been incredibly
01:14
volatile and as a newly publicly traded
01:16
company wejo stock has certainly felt
01:18
that impact so given the harsh market
01:21
conditions of today why should investors
01:23
believe that wijo is a good investment
01:25
to answer that question i sat down for
01:27
another exclusive interview with richard
01:29
and we caught up on everything that we
01:31
joe has been up to including some big
01:33
developments that they're working on
01:35
with palantir and microsoft if you watch
01:37
until the end you'll gain a solid
01:38
understanding of connected vehicle data
01:40
and you'll also learn about we just
01:42
newest offerings with palantir their
01:44
other avenues for growth and the future
01:46
of connected mobility full disclosure
01:48
i'm not a financial advisor and nothing
01:50
in this episode is financial advice i
01:52
took no compensation for this interview
01:54
but i am currently a ouijo shareholder
01:56
and may increase my position in wijo in
01:58
the near future based on this research
02:00
your time is valuable so i've cut out
02:02
all of the fluff and timestamps are
02:04
enabled for your convenience so let's
02:06
get right into it richard it's great to
02:08
see you the last time we spoke was right
02:10
before your reverse merger with virtuoso
02:12
acquisition corp you know the markets
02:14
have gone wild this year how have you
02:16
been how's wejo been it's been an
02:18
interesting start we're like any tech
02:20
stock i think one thing i'm really happy
02:21
about and we're really excited about is
02:22
we've got a good trading volume we'd all
02:24
love for the share performance to be
02:27
better but we are reflective of of the
02:30
growth tech tech stocks more broadly
02:33
we've had some great news flows since we
02:35
last met so you you'll have seen the uh
02:37
palance evos announcement last week we
02:40
made our announcement with neuroledger
02:41
microsoft at the beginning of the year
02:43
at ces we made an announcement of our q3
02:46
results in december we gave a business
02:48
update where we announced our first uh
02:50
more than a million dollar contract
02:51
where we licensed software to a motor
02:54
manufacturer in the us
02:55
so it's been an interesting three months
02:58
really excited this year we're excited
02:59
about some announcements we're going to
03:00
make
03:01
really excited by the support we've had
03:03
from your from from some of your
03:05
followers um yeah it's been good
03:08
that's awesome yeah i'm really excited
03:09
to get into it with you over the course
03:11
of this interview i can't wait looking
03:12
forward to it for the investors who are
03:14
watching now who didn't catch the most
03:16
recent interview with wijo you work with
03:18
connected vehicle data what does that
03:20
mean and what opportunities is uijo
03:22
chasing in the market exactly well we
03:24
work with both autonomous vehicles
03:26
electric vehicles and connected vehicles
03:28
and our overall sort of pool of data we
03:31
have we have about 12 million vehicles
03:32
live on platform what what a connected
03:35
vehicle is is which which encompasses av
03:37
and ev as well is that the vast majority
03:40
of new vehicles sold globally every year
03:43
have embedded connectivity they have
03:45
they have communication devices called
03:47
the tcu
03:48
embedded in fabric or the network of the
03:51
car whether the car's a modern car
03:53
running something called an ethernet or
03:55
whether it's running a more more
03:56
historic network called can these
03:59
networks in the car are a circular
04:00
network and every sensor in the car so
04:04
you know there can be a every button in
04:06
the car
04:07
there might be a sensor in the window
04:09
detecting if something's stopping the
04:10
window closed
04:12
there'll be a sensor in the in the
04:13
windscreen in the windshield wiper
04:15
there'll be a sensor measuring
04:16
temperature outside
04:18
you catch my drift there are thousands
04:20
of sensors in the car
04:22
um typically a seat can have up to 35
04:25
sensors in it so there can be sensors on
04:27
the buttons of the seat
04:29
so what we do is we work with motor
04:31
manufacturers industry term oems
04:34
and we receive access to up to 220
04:37
sensors in vehicle
04:39
so we get the live location of vehicles
04:41
with consent of the driver and we see
04:44
seven set of vehicles driving around new
04:45
york for example we see 20 of all
04:47
vehicles driving around detroit you can
04:49
imagine wise you know when we're working
04:50
with some of the world's largest oems
04:52
where they have their headquarters in
04:54
detroit that will explain why we have 20
04:55
of all vehicles on the road but it's
04:57
quite surreal walking down the streets
04:59
of detroit and seeing one in five
05:01
vehicles and i'm thinking when i walk
05:03
down there
05:04
i'm getting from that vehicle from that
05:06
vehicle from that vehicle so we get live
05:08
location and that's helping smart cities
05:10
understand about emissions and cities
05:12
understand about the weight of vehicles
05:13
and how that can affect how it can
05:15
affect the the the deconstruction of
05:17
roads but then we're also getting
05:18
powertrain data so we know uh engine
05:21
temperature
05:22
i previously had a vehicle where i
05:24
actually sent the ceo uh a spreadsheet
05:26
tracking the temperature raising rising
05:28
on my on my engine and i i'm a geek
05:32
and i wanted to understand why uh why i
05:34
was losing so much oil in the vehicle
05:36
and i showed in a 20-minute period this
05:38
rising of temperature
05:40
and that's that's sort of the example of
05:41
some of the live data from vehicles you
05:43
know so we get powertrain we get battery
05:45
status
05:46
and later on we can talk about palantir
05:48
and our announcement we're there with
05:49
what we call evos but we're getting live
05:52
battery data from from vehicles so we
05:55
know how batteries are performing we
05:56
know we know when a vehicle's parked we
05:58
know he's got in park we know if it's
05:59
then being charged or not so we've got a
06:01
huge amount of data and we process over
06:03
17 billion data points a day we process
06:06
over 13 trillion today
06:08
so it's a broad array so if you think
06:10
about any button in the car any any any
06:12
moving anything mechanical in the
06:14
vehicle anything involves the vehicle
06:16
moving there'll be a sensor and that
06:17
data is potentially available for
06:19
analysis which we receive wedo just
06:21
finished their first quarter as a
06:23
publicly traded company what makes wejo
06:25
stand out in today's tough market
06:27
against other young companies that
06:30
investors should be thinking about right
06:31
now we are the in terms of live vehicles
06:34
on platform in terms of scale scale
06:36
revenues we are the market legend from
06:38
connected vehicle av and ev data
06:41
um we built incredible relationships
06:43
with likes of microsoft who don't tend
06:45
to do public equity investments we've
06:48
we've built an incredible relationship
06:49
with palantir
06:51
and they've been not only a great
06:53
investor but also a great commercial
06:55
partner and also we've had we have the
06:57
private backing or private company
06:59
backing from general motors for three
07:00
years we've continued to backers in the
07:02
public market as well where they're also
07:04
one of our pipe investors so we feel
07:06
with standout in terms of alternative uh
07:09
data data providers in in the connected
07:11
vehicle data space
07:13
in all key metrics
07:14
we're transparent about uni economics um
07:18
the and and then beyond that
07:20
what we found with the um being in the
07:22
public markets is that the inbound
07:23
demand is just has has been even greater
07:25
we're getting more and more demand for
07:27
more and more different types of
07:28
agencies different industries you want
07:30
to do more with with the data that
07:33
we control and represent i read your
07:35
investor presentations and so on and it
07:37
seems that one of the things that you're
07:38
getting into now is smart infrastructure
07:41
do you want to talk a little bit about
07:42
that as well as maybe your relationship
07:44
with palantir on this latest product
07:46
smart infrastructure is a broad
07:48
definition for industry but it's this
07:49
idea of having
07:51
more smarter data sources we just
07:53
announced with palantir something called
07:55
evos the ev operating system and it's
07:58
come from really that that that i was
08:00
working with until we was working with
08:02
palantir
08:03
and we we we so we've seen this myth
08:06
about about the industry it has around
08:08
range anxiety everyone worries about the
08:10
size of batteries in the vehicle
08:12
that's not the challenge the challenge
08:14
is is that is actually you know a
08:15
typical driver drives 40-50 miles a day
08:18
most cvs all evs now have more than
08:21
enough battery life to do multiples of
08:24
those journeys it on a day-to-day basis
08:27
so the challenge really is is that if
08:29
the vehicle can be plugged in at home or
08:31
can be plugged in where the vehicle
08:32
whether it was parked at night
08:34
then how can you actually provide
08:36
real-time management for the grid that
08:38
not all vehicles are sat trying to
08:40
charge overnight at the same time how
08:42
can you throttle that if a vehicle is
08:43
plugged in from saying
08:45
9 p.m to
08:46
7 a.m how can you throttle that a group
08:49
of vehicles and agree to charge from
08:51
9 00 pm to 10 p.m then another group are
08:53
going from 10 p.m to 11 p.m
08:56
if you can work with the grid and that's
08:57
what the evos is doing with palantir is
09:00
that we're providing a real-time dynamic
09:01
view so a grid then knows when when
09:04
actually they they they need to direct
09:06
more power to a particular area but
09:08
conversely we can then work with oems
09:10
who can then just you can then you can
09:11
then build apps within the car or the
09:13
vehicle that isn't just taking charge
09:14
all the time but it's always taking
09:16
charge it's plugged in but taking charge
09:18
when the grid can handle it that's going
09:20
to fundamentally reduce a lot of the
09:22
infrastructure budgets that have been
09:23
asked for the moment which is assuming
09:25
that everyone needs to be upgraded it
09:26
doesn't that's not the challenge the
09:28
challenge is to control the power to the
09:30
vehicles at the right time which doesn't
09:32
convenience the consumer doesn't mean
09:34
his inconvenience at the driver that's
09:36
what's exciting to me about evos is this
09:38
idea that that we can now help
09:40
power energy companies make more
09:43
intelligent decisions and truly give
09:45
some smart infrastructure sure that
09:47
makes a lot of sense and so the idea
09:49
there is to really optimize ev
09:51
infrastructure investment is that like a
09:53
fair way to
09:54
yeah and you know if you look at the
09:56
biden infrastructure act um there's
09:58
about eight billion allocated for ev
10:00
infrastructure so it's helped that you
10:02
know and that's and that's not a lot
10:04
really in terms of the spend
10:06
but actually we can help optimize how
10:08
that is spent how that's effectively
10:10
spent i get to get to get the best
10:12
coverage so that all consumers who want
10:14
to own a vv or drive drive and ev can
10:16
actually have an ev that's always on
10:18
demand always readily always charged
10:21
they don't need to worry about the fact
10:22
it was only charged from 2 a.m to 2 30
10:24
a.m
10:25
it's got enough charge to do their
10:27
typical journeys
10:28
they can always override they can always
10:29
say they want to the full charge but
10:31
actually most people don't need that in
10:32
their vehicle that's so interesting yeah
10:34
so it it sounds like this os with
10:37
palantir really focuses on ev
10:40
infrastructure optimization do you plan
10:42
on launching other os's that might uh
10:45
optimize for other things for example
10:47
infrastructure for traffic management
10:49
optimizing traffic lights and stop signs
10:51
and like things like that is there plans
10:53
for spin-offs for other versions of
10:56
optimization os's with palantir so in
10:58
terms of we just studio and we just
11:00
studios is is is now becoming more
11:02
powered by by foundry uh we built studio
11:05
before before before foundry and we're
11:07
now doing a a migration across and
11:09
leveraging all the power that foundry
11:11
gives to us we've built products now
11:13
where we know uh when traffic lights are
11:15
are red or green sometimes that's from
11:18
vehicles identifying that or determining
11:20
that the the traffic lights change from
11:22
red to green sometimes it's taking
11:23
third-party data sources in so and
11:25
that's another example of of how we can
11:27
how we how we can help smart city
11:29
infrastructure
11:30
but going beyond that
11:32
what we're now designing we announced at
11:34
the beginning of january is what we is
11:35
what we call our digital twin although
11:37
although palantiy gave it away and said
11:39
actually it's we're calling it meta twin
11:41
we see 77 million car journeys a day or
11:44
vehicle journeys a day in cities in the
11:45
us for example so we're seeing huge
11:48
amounts of data we've collected over 507
11:51
billion miles of data we can now
11:53
recreate that data and create a 3d
11:56
environment a a meta twin a digital twin
11:58
where you can see all those vehicles
12:00
driving around
12:01
but then the question leads us to a city
12:03
planner well what would happen if a road
12:05
was changed from two lanes to one lane
12:07
and actually i was talking to a
12:08
portfolio manager and that was his exact
12:10
question he said hey i live in austin
12:12
and on my road we we lobbied the city
12:14
planners go from two lanes to one lane
12:16
did that help mice did that help the the
12:18
congestion on my on my road and we could
12:20
actually show with wijo's studio in 2d
12:23
right now the benefits to an easier
12:26
flowing of traffic well now in in our 3d
12:29
environment which will be launching this
12:30
year you'll be a smart city planner will
12:33
be able to reconfigure a road and
12:34
actually see in real time vehicles
12:36
vehicles changing their behaviors
12:38
because we've learned from more than 11
12:40
million instances what happens when two
12:42
vehicles come together and we've had
12:43
data from both vehicles so it's really
12:45
exciting and it's hugely powerful and
12:47
shows the power of of of the data we've
12:49
collected and the power of what
12:51
foundry's not giving to us as a business
12:54
so that digital twin of both cars and
12:56
cities is something that's enabled by
12:58
foundry
12:59
they the first first alpha we built
13:02
in-house and now we're working through a
13:04
foundry to now we're now migrating all
13:06
that into foundry so we are in active
13:08
conversations planning that through
13:11
we're actively moving wijo's studio
13:13
within the foundry environment so
13:15
foundry so we built studio with it
13:17
within our own innovation lab and then
13:19
now we're migrating to foundry now we're
13:21
leveraging what foundry can now give to
13:22
us at scale so it's so we we sort of
13:25
innovate in house and then we work with
13:27
park with palantir to them to then super
13:29
scale out so yeah it's a great
13:30
partnership that was with with the uh
13:33
guys that uh guys and girls at palanti
13:35
all day yesterday in london and we you
13:36
know we've got some really exciting
13:37
stuff where we'll be doing this here
13:39
there's like so much to dig into here
13:42
digital twins and talking about things
13:44
like the metaverse uh have been on
13:46
everybody's mind lately but i think the
13:48
practical applications here are
13:50
something worth digging into a little
13:51
bit so you're building digital twins on
13:53
both the individual car level and a
13:56
hypothetical city level is that correct
13:59
every 3d game is a matter of some sort
14:02
so this is just another matter uh it'll
14:04
be a match environment for for
14:07
scientists for r d labs to understand
14:09
how vehicles behave how how they how
14:12
they will change with differing dynamics
14:15
it's great for the autonomous vehicle uh
14:17
industry community for example
14:20
um there's a there's over 40 dmv
14:22
licenses for oems who are doing av
14:24
trials in us
14:25
but you'll notice a theme they're all
14:27
focused in the same city
14:29
they're all folks in very small parts us
14:31
you know apart from tesla there's no
14:33
national approach for oems about how
14:36
they're going to make their cars make
14:38
their avs be performant for more than if
14:41
a few vehicles
14:42
so we will open and we will open our
14:44
meta twin that any oem with av can run
14:48
their own proprietary av algorithm
14:50
within our twin and understand about
14:52
what happens when they come across a
14:54
human driving a vehicle
14:56
you know we understand the outcomes
14:57
already of a human driving at 55 miles
15:00
an hour where the temperature is 2
15:02
degrees outside it's
15:04
2 30 a.m in the morning it's snowing we
15:08
know what happens next if a vehicle hits
15:10
another vehicle we've already pre-built
15:12
that or machine learning and we've built
15:14
this dictionary of understandings so
15:16
it's hugely powerful and that'd be a
15:18
great way
15:19
of bringing forward the the av
15:21
revolution
15:22
the evolution there's a lot of
15:25
but they're limited to san francisco
15:27
they're limited to arizona you know
15:28
there's the the the av community now
15:31
needs to know about how they can bring
15:32
all that bring forward and enable all
15:35
communities to have better access to all
15:36
these vehicles that'll only work in a
15:39
digital environment to start with
15:40
sure no that makes a lot of sense
15:42
so last year tesla talked a lot about
15:45
how one of their big reasons for going
15:47
to vision only was to be able to build a
15:49
simulation engine and be able to learn
15:52
from you know computer simulations video
15:54
game graphics you know this digital twin
15:57
of basically america and pull visual
16:00
data from that simulation space that
16:02
digital space back into the physical
16:04
space do you have any plans like for a
16:06
similar capability where now that you've
16:08
built this digital twin you can simulate
16:11
data and integrate that with your real
16:13
data streams or add value by simulated
16:16
data or is this strictly for
16:18
experimentation no i mean we we we want
16:20
to get to a point where any oem can log
16:23
in and they can go to any city
16:25
and they can they could they can put
16:27
their own defined algorithm of how their
16:28
av drives
16:30
and they can then drive around they can
16:32
and we and we already have videos
16:35
where which we're building a house where
16:37
we where we can see what's going on
16:39
what's happening next you know so we and
16:41
we we simulate
16:43
pedestrians crossing the road we
16:45
simulate bikes going down the road this
16:48
is all this is all built it's in alpha
16:50
we are not selling a date when we're
16:51
launching this
16:52
but this is our plan and this is why
16:54
when when we when we announced we are
16:56
going to market at the end of last year
16:58
we were very clear we are becoming the
17:00
communication and data stack of industry
17:03
we've collected a huge amount of data
17:05
we're working with multiple oems they're
17:08
trusting us with their data and we're
17:09
now showing the power of that trust
17:11
we're now showing and building some
17:12
incredible products they're going to
17:14
revolutionize smart cities make the av
17:16
revolution happen
17:18
make ev more feasible quicker rather
17:21
than the worry of range anxiety and this
17:23
is these these are some examples of what
17:25
we call data for good you know we're
17:26
thinking about where data adds value not
17:29
just cynically cell data but actually
17:31
how does data add value how does it
17:32
reduce the cost of your insurance how
17:34
does it reduce the cost of repair of
17:36
repairing the vehicle
17:38
as in other examples
17:40
no those are great yeah and so i think
17:43
that makes a lot of sense your core
17:45
mission with why you've partnered with
17:47
palantir who works with novel data
17:49
companies to enable a lot of these
17:51
things to happen right the amount of
17:53
data integration aggregation processing
17:57
and insights you generate require
17:59
something like a foundry to spin up in
18:00
less than a couple decades right can you
18:03
talk a little more about oh go ahead
18:04
please no i was going to i mean i i i
18:07
saw uh yesterday at pounds here where we
18:09
were giving some demonstrations
18:11
and what they did with the uh the uk
18:14
health service called the nhs
18:16
they integrated over 150 nhs trusts they
18:19
integrated hundreds of factories or
18:22
warehouses in in days to join during the
18:25
kovic pandemic and gave this real-time
18:28
view of how ppe was being distributed of
18:30
how of how of how and then eventually
18:32
how vaccines were being distributed and
18:34
it all then and it was only simplified
18:36
to us to a dashboard that the nhs ceos
18:39
could see the government could see and
18:41
it was a great example of actually data
18:43
for good for palantir the great example
18:45
of how foundry can can be can be super
18:47
scared to be skilled so quickly
18:49
in terms of why palin's and ouija work
18:52
together we've now got the same approach
18:54
we're work we work with oems and we have
18:56
access to their data we're now working
18:58
with what they call the tier one and
18:59
tier two supply chain the tier one tt
19:02
supply chain are the parts providers
19:04
that go into your vehicle whether it's
19:06
an engine manufacturer whether it's a
19:07
seat manufacturer you will find an oem
19:10
has hundreds of suppliers in their
19:12
supply chain
19:13
what the oems have never done is that
19:15
they've never shared
19:17
insights of the data from sensors back
19:18
with their supply chain after vehicles
19:20
have been sold that's where we're
19:22
working with foundry is that we have
19:23
this ambition
19:24
that the hundreds and thousands of tier
19:27
one tier two tier threes in the overall
19:29
oem supply chain will all be integrated
19:32
into foundry
19:33
and that's our ambition is and that's
19:35
what and that's and that's you know
19:36
we're applying the same idea that
19:38
parents have found you with with the
19:40
with the aerospace industry with with um
19:42
with airbus we're now doing the same
19:44
with them in the automotive space
19:46
and you know it's it's been an
19:47
incredible partnership we've uh you've
19:50
hopefully seen that we've done some
19:51
joint joint marketing or already with
19:53
likes of evos we're gonna have more to
19:56
be shouting out this year next year and
19:57
beyond
19:58
that's super exciting yeah so just just
20:01
to bring the wijo and palantir section
20:04
uh to a cohesive conclusion the way i
20:06
kind of imagined this building out is
20:09
you and palantir together are building a
20:11
digital twin at the macro level and at
20:14
the car level people are able to log in
20:17
and see everything going on you know car
20:20
by car but also this macro level traffic
20:23
you know what's going on in the city
20:24
let's change this road let's look at
20:26
what happens if we invest here and how
20:28
that affects everything going on as well
20:30
as make decisions using that digital
20:32
twin in near real time is that fair like
20:35
one of the things that we joe often
20:37
talks about is being the communication
20:39
wear between vehicles is that vehicle to
20:42
vehicle is that vehicle to cloud to
20:44
vehicle is that vehicle to digital twin
20:46
to vehicle where does the decision
20:48
making actually take place and how does
20:51
that work at like a detailed data level
20:53
so the capsule is v2x so it's vehicle to
20:55
anything so whether it's whether it's
20:58
vehicle to vehicle whether it's vehicle
20:59
to cloud we're we're working and we we
21:02
can talk in a second about what we're
21:03
doing at the edge with terms of vehicles
21:05
and
21:06
edge um but it's vehicle to anything
21:09
it's this idea that the vehicle
21:11
vehicles um today have not been
21:14
have not had um ai or or ml sort of
21:18
focus around around data leaving
21:20
it's been quite a a it's been it's been
21:23
quite sort of formulated in terms of
21:25
what data should or should leave the car
21:27
in our vehicle in what sort of speed
21:29
or and
21:30
which data points it is we then provide
21:33
the intelligence at the cloud we're now
21:34
empowering vehicles to actually have
21:36
more that power more of that
21:37
intelligence back in the vehicle but but
21:39
ultimately right now we enable vehicle
21:41
vehicle to vehicles anything sharing of
21:43
data so vehicle to anything does that
21:45
mean literally things like a vehicle can
21:48
if a vehicle is waiting at an
21:50
intersection it's three in the morning
21:52
there's no one coming and a light is red
21:53
the vehicle can connect or communicate
21:56
to the traffic light and say hey dude
21:58
turn green i'm the only one here or what
22:00
exactly does it mean vehicle to anything
22:03
yeah so vehicle to infrastructure is is
22:05
um is is is exactly that it's this idea
22:08
that you know
22:10
we're already getting traffic light data
22:11
whether it's traffic light data from the
22:13
traffic lights themselves or whether
22:14
it's inferred from other vehicles this
22:16
idea of a more effective infrastructure
22:19
uh and communication to infrastructure
22:21
is is part of our remit it ultimately
22:24
then leads to better safety reduce
22:27
emissions on reduced emissions and
22:29
emissions is broad definition it's not
22:31
just think about combustion engines it's
22:33
all attributes for vic which could then
22:34
lead to emissions
22:36
so one of the infrastructure use cases
22:38
i'm often thinking about are which
22:40
chargers and which gas stations have
22:42
lines right that's not something we
22:43
really get to see today is that
22:45
something that emerges from being able
22:47
to collect all this data what charges
22:49
are open where it makes sense to put
22:50
chargers where it makes sense to you
22:52
know load up on gas is it more about
22:54
path planning is it more about
22:56
infrastructure where do you see
22:58
like the big picture net effect
23:00
happening so actually we've already got
23:02
winning running in in in our foundry uh
23:05
in our foundry installation at ouijo so
23:07
you can already see live dodging
23:08
availability with it within within
23:09
wijo's foundry environment so we're
23:11
already getting
23:12
data from both vehicles and data from
23:14
infrastructure so that's a great example
23:16
of we know when infrastructure is
23:18
failing we know when charges are failing
23:20
that that has got to be addressed as
23:22
well but we're already seeing that and
23:23
that's a great example of vehicle to in
23:25
vehicle to infrastructure communication
23:27
but infrastructure can be street lights
23:29
if there's no vehicle movements on the
23:30
street why do the lights need to be on
23:33
you know that that can say
23:35
so there's all sorts of benefits that go
23:37
from vehicles to infrastructure vehicle
23:39
to infrastructure can be a parking lot
23:42
can be curbside parking
23:44
anything that ultimately can lead to
23:46
more efficient use of a vehicle which
23:48
then always leads to better safety and
23:51
less congestion
23:52
anything that we that we class as day to
23:54
be good we we we are focusing on we're
23:56
thinking about as a business
23:58
so with all that data being communicated
24:01
between cars between cars and
24:02
infrastructure and between you know the
24:05
physical wear and the digital twin comes
24:07
a lot of issues with transmission
24:10
latency storing all that data how do you
24:12
guys handle that
24:14
so we so we process over 17 billion data
24:17
points a day at peak 500 000 data points
24:20
a second
24:21
um
24:22
right back in the cloud we we hold over
24:24
eight petabytes of data so you know so a
24:26
big amount of data
24:28
um and there's all sorts of challenges
24:30
around that so we we're continually
24:32
machine learning from that data asset
24:35
um so we needed a cloud partner that
24:37
could help us
24:38
help us be able to break that data down
24:40
and and microsoft azure has been a great
24:42
has become a great partner of ours
24:44
who've helped us
24:46
be better with data that's not to say we
24:49
don't have google bigquery instances
24:50
that's not to say we don't have aws s3
24:53
instances but when we but when we were
24:54
looking at the next five years
24:56
we looked we looked at the whole
24:58
industry of cloud partners and we and we
25:01
recognize microsoft as because is doing
25:03
really interesting things in automotive
25:05
and has built some really interesting
25:07
products that were very relevant to what
25:09
we were wanting to do
25:10
one of the biggest challenges in
25:11
industry and i mentioned to you
25:13
mentioned earlier on this podcast is
25:15
that vehicles right now haven't i'm not
25:17
we're not that intelligent about
25:19
determining when data should be sent
25:21
and there is a bottleneck the bottleneck
25:23
is
25:24
the 4g network or becoming the 5g
25:26
network
25:27
so and that and that bottleneck means is
25:29
there any there's only so much data that
25:30
can leave the car
25:32
and the high value the higher value data
25:35
the data that can help save the av
25:37
industry
25:38
um such as image data is is will use of
25:41
a big proportion of that of of this of
25:44
the size of the pipe or the bottleneck
25:45
that the communication channel
25:48
so
25:49
we're we we've already been faced with a
25:51
challenge of
25:52
we're here we're hitting a sort of a
25:54
critical amount of data leaving vehicles
25:56
and that means we don't need all the
25:57
data all the time and yet we can't
25:59
instruct a vehicle in real time to stop
26:00
sending us data so we end up taking more
26:02
and more data and even if we don't even
26:04
if we learn no get no more value from
26:06
that particular data field or data point
26:09
so we've so we've been working for the
26:10
last two years on how we can have
26:13
machine learning outcomes or when data
26:15
should or should or shouldn't be sent
26:16
from the vehicle
26:18
we've proven we can reduce the data
26:19
overhead by 80
26:21
and still have
26:23
a 100 percent intact data point at the
26:25
other end so by reducing the data
26:28
overhead by 80 that means you can have
26:29
more data out to the vehicle
26:31
which is going to be absolutely perfect
26:33
for av
26:34
the i av has not yet finished learning
26:37
everything needs to learn the whole
26:38
industry hasn't so we've developed a
26:41
developed approach where at the vehicle
26:42
level we can we can we can throttle the
26:45
data when it needs to leave the vehicle
26:46
leave the data so the dates easily
26:48
vehicle or not then on the cloud side
26:50
we're now working with microsoft and
26:51
we've learned we've launched a product
26:53
called neural edge where we're
26:55
leveraging edge-based expertise from
26:57
from the azure solar portfolio to
26:59
rebuild the data
27:01
and the idea that we can rebuild the
27:03
data and it's still being 100 in in
27:05
integrity
27:06
means that we can then have a much
27:08
bigger
27:09
data pool to then do even more enhanced
27:11
machine learning and then back to our
27:13
digital twin scenario earlier that's
27:15
going to be so important because one day
27:17
what we want is the 40 evs with dmv
27:20
licenses to have to to do av we want
27:22
every one of those different avs be
27:24
driving within our digital twin and then
27:26
they all learn together how they can
27:27
interact together in this new world
27:29
where there's going to be no one
27:31
proprietary av approach there's going to
27:33
be tens or potentially hundreds and
27:36
these vehicles are all going to need to
27:37
live together and they get there's going
27:39
to need to be a better way of handling
27:40
data rather than at the moment when a
27:42
navy is machine learning how to drive
27:44
around a city
27:45
the data stays within the vehicle all
27:46
day and then it's downloaded at the end
27:48
of the day well that can't happen when
27:49
you've got millions of vehicles driving
27:51
around there needs to be a real-time way
27:52
for data to be shared and that's what
27:54
we've been building is what we call
27:55
neural edge but this idea that we will
27:57
become the comstat that enables whether
28:00
it's evs or avs or connected vehicles a
28:02
real-time way a sub-millisecond a
28:05
low-latency way of dating leaving a
28:08
vehicle and going back to the vehicle
28:10
is that so is this just about you know
28:12
vehicles with four wheels on roads or
28:14
can this be expanded to flying vehicles
28:16
like evitals scooters you know urban
28:19
scooters in cities talk to me a little
28:21
bit about all of the other edge cases
28:23
that this can or can't expand to i mean
28:26
any any mobility um so micro mobility um
28:30
and you know i was looking at the
28:31
jackson one the other day which is a
28:33
single amanda uh craft flying craft yeah
28:37
anything that enables mobility anything
28:39
that's got smart mobility
28:41
we should be able to take data from it
28:44
in the in the new world of mobility
28:46
micro mobility or scooters will need to
28:49
know when vehicles are closed there
28:51
needs to be a much more effective a much
28:52
more efficient way of sharing data
28:54
between all types of mobility devices
28:57
vehicles sure and so one of the implicit
28:59
things that i think we haven't said
29:02
explicitly is this also works for
29:04
autonomous logistics right so we've
29:06
talked a lot about you know scooters
29:08
cars and then flying vehicles charging
29:11
infrastructure these are all things that
29:13
people tend to touch but you know
29:15
another thing that's really important to
29:16
optimize is you know trucks leaving
29:19
warehouses and getting to you know big
29:21
businesses right loading and unloading
29:24
between ports and other things like that
29:26
where how does wijo help you know
29:28
industrial players not just
29:30
uh you know citizens with
29:33
point a to b transportation yeah i mean
29:35
we've got we've got so many ideas we're
29:38
we're focusing on on our on our core
29:41
which is vehicles um absolutely future
29:44
i'm huge excited by say what bright drop
29:47
are doing um huge excited about about
29:49
about about the robotics that uh firms
29:52
like academia are doing um there's all
29:54
sorts of different ways of efficiency
29:57
around around logistics
29:59
for us we're stopping the vehicle level
30:01
for now whether it's micro mobility or
30:03
or whether it's trucks we won't go into
30:05
logistics but you but in terms of
30:07
platform we're building
30:09
it's it is it is it is like foundries
30:13
it's compatible with with almost any
30:14
data source sure
30:16
but right now our focus is vehicles you
30:18
know we are a connected vehicle an av
30:20
and a an a an autonomous vehicle and an
30:23
electric vehicle business got it no that
30:25
makes a lot of sense and it's great to
30:26
clarify you know what we joe does and
30:28
doesn't touch so we can you know better
30:30
understand the scope the unit economics
30:33
um so talk to me a little bit more about
30:35
the via the communications right so this
30:38
neural edge program sounds like it's
30:40
more than just you know toning down so
30:42
you only get one out of every five
30:44
points out of the vehicle and then
30:46
reconstructing it back on the cloud is
30:48
there any other like edge computing
30:50
going on at the vehicle level what else
30:51
does neural edge provide is it security
30:54
and standardization what else beyond you
30:56
know interpolation does neural edge do
31:00
it won't do standardization that's i
31:02
mean that that's there there are a lot
31:04
of tier one semiconductor providers who
31:07
have their own approach to to how data
31:10
is is is is um is is around the vehicle
31:14
so we we we're not doing that we're
31:16
working with semiconductor providers
31:18
we're working with the tier ones who
31:19
have their own approach about how they
31:21
want to have their own their own their
31:23
own data formats
31:24
we're providing a platform where we can
31:27
see the data in whatever format it is we
31:29
run something called a common data model
31:31
where we then standardize at the edge so
31:34
rather than actually forcing a standard
31:36
on anyone else we say
31:38
like foundry does we can take the data
31:40
in any format
31:41
let letters have any format at the edge
31:44
of the vehicle
31:45
let us use our our own proprietary
31:47
machine learning that when the data
31:48
showed or shouldn't leave the vehicle
31:50
will determine whether it needs to be
31:51
sent on a sub late and some latent base
31:53
a sub millisecond latent basis or
31:55
whether it should be set at the end of
31:56
journey at the end of the journey
31:58
there's not the necessity for the
31:59
real-time selling of data so that's the
32:01
intelligence at the edge that we that
32:03
we've been developing
32:04
and then and then and then when it's
32:06
rebuilt the other end if we've only
32:08
determined to receive some data now
32:11
um but we see other data in a journey we
32:13
need to rebuild that we need to make
32:14
sure that the integrity is still there
32:16
of the data the data for for further
32:18
machine learning in the future that's
32:20
that's our focus our focus is the
32:22
efficiency of data leaving the vehicle
32:24
where we're clear what we do we're clear
32:27
we're about being a data ecosystem and
32:28
com stack for industry
32:30
we we're that's our swim lane that's our
32:33
focus and our focus is ultimately about
32:35
about vehicles becoming more and more
32:36
efficient to ultimately drive to better
32:39
safety in the roads low emissions and
32:41
and less congestion uh it sounds to me
32:44
like what neural edge does is it enables
32:47
you to be able to say hey tone down the
32:49
amount of data we send in real time so
32:51
that we can reconstruct it on the cloud
32:53
you know reconstruct this incomplete
32:55
data set on the cloud and get all of the
32:57
data without having to transmit it is
33:00
that true
33:01
yeah so earlier when i was talking about
33:03
about the fact that the oems most
33:05
manufacturers are all wondering about
33:06
how they can get how they can get av to
33:07
scale
33:08
and one of the challenges they've got is
33:11
that is that at the moment in terms of
33:13
their their lab environment or their rnd
33:16
environment they have
33:17
maybe hundreds of cars driving around
33:19
cities if that
33:20
less than hundreds usually
33:22
ultimately
33:24
they're collecting huge amounts of data
33:25
in vehicle that they don't send back in
33:28
real time
33:29
they don't they can't do real-time views
33:31
they may they may all be machine
33:33
learning with it within the vehicle for
33:34
that driving today then they'll review
33:36
the vehicle they'll review the data
33:38
after the event
33:40
that's not scalable
33:41
if oems want to scale to millions of avs
33:44
they need to have a way
33:46
that of actually doing more real-time
33:48
learning of data backing cloud in a
33:50
real-time environment so what neural
33:52
edge does is it enables the real-time
33:54
sending of data and the real-time
33:56
processing of data at the edge
33:58
where we can then where we can then
34:01
share our own learnings
34:03
which we've had from other oems share
34:06
our share our own machine learning
34:07
outcomes provide access to our own
34:09
digital twin and that will enable avs to
34:12
become mass market rather than this ad
34:14
hoc approach to every vehicle doing its
34:16
own
34:17
storing its own data before after event
34:19
analysis
34:21
yeah you know one thing i just realized
34:22
is like usually what i think of is
34:25
palantir is sort of a huge player in
34:28
edge computing but you're leveraging a
34:30
lot of microsoft technology to make that
34:32
happen can you describe a little bit
34:34
more about you know when we think about
34:36
palantir and microsoft i naively think
34:39
about them as competitors but it seems
34:41
ouijo has a great working relationship
34:43
with both is it like
34:45
microsoft is your partner for computing
34:47
and palantir is your partner for data
34:50
analytics like how how have you managed
34:52
to bring these two seeming competitors
34:54
together and what
34:56
can you clarify their roles like for
34:59
wijo
35:00
no so you you can actually run uh
35:02
pounce's foundry with it within varying
35:04
cloud environments so we chose we chose
35:06
to up to what to run it within the user
35:08
environment um there's no competitive
35:10
issue there uh you need to choose a
35:12
cloud vendor you need to choose a vendor
35:14
that that that that provides you a scale
35:17
of of being able to process huge volumes
35:19
of data
35:20
so
35:21
there is not a competitive tension um
35:24
and i'm not i'm not something that that
35:26
we've experienced um pounds has been a
35:29
great partner microsoft's happy with our
35:31
relationship as well so it's been a
35:33
great triumvirate that we're leveraging
35:35
the best of foundry's distribution it's
35:37
best in terms of data analytics and
35:39
we're using the best of azure with the
35:41
ability to real-time process data and to
35:44
build and to build other tools with with
35:46
within within the user environment which
35:48
don't leverage foundry so for example
35:50
we've built a real-time consent engine
35:53
for um for for an oem or and privacy
35:56
engine
35:56
um and we've licensed that on that and
35:58
that was actually our first more than
35:59
million dollar contract where where the
36:02
oml now leverages our our expertise
36:04
within the user environment to anonymize
36:07
data in real time for millions of
36:09
vehicles so and that's not something
36:11
that that that you would do in the
36:12
foundry environment it's certainly
36:13
something you do within the zero
36:15
environment that's really exciting and
36:16
it sounds like you're getting both
36:18
low-level data like individual sensors
36:20
and high-level data enough to build
36:21
digital twins of not just vehicles but
36:23
all the way to smart cities is that
36:25
correct
36:26
yeah and you know it's really
36:27
interesting um you know we talk about
36:29
low level sensors and uh one of my new
36:31
one of my new favorite sort of case
36:33
studies is that we're getting there's
36:35
there's microphones in the wheel arches
36:36
of certain cars which are the low level
36:39
is there to detect a vehicle aquaplaning
36:41
so when when when the wheel is going
36:43
through rain and then it stops turning
36:45
so the vehicle is losing control
36:47
so the vehicle so the sensor was was was
36:49
was used to to measure this vehicle
36:52
aquaplaning and then and then the abs
36:54
the esp can be can be activated
36:56
accordingly
36:57
well we've repurposed so to speak this
36:59
this low level sensor
37:01
and now we're and now we're identifying
37:02
frequencies of so we can identify panel
37:04
damage in vehicles so we can identify
37:07
panel damage within 12 centimeters
37:10
so what four inches
37:12
um so we can identify within about four
37:14
inches where a panel has been damaged so
37:15
you think about the use case of
37:16
insurance
37:18
if an insurer or a carrier can can be
37:20
informed in real time where the panel
37:22
damage has occurred
37:23
before the vehicle's even even being
37:25
picked up tends to take for body shop
37:27
the insurer can already make sure the
37:28
parts are on order or if the body shop
37:30
is is prepped to repair say aluminum uh
37:34
compared to say um
37:36
plastic there's another example
37:38
so then and then you think about smart
37:40
cities um we
37:42
as i mentioned to you before we see 20
37:44
percent of vehicles moving around
37:45
detroit seven percent in in new york
37:48
and another cities including actually
37:49
six percent of the whole state of
37:51
california so having this
37:53
live data
37:54
in real time means that we see really
37:57
interesting things you know we identify
37:58
in real time where an intersection stops
38:01
stops taking traffic
38:03
that might seem obvious but we're
38:04
working with say one of the world's
38:06
largest e-commerce companies where
38:08
they've told us we've improved delivery
38:09
routing by 30
38:11
which means they can deliver more
38:12
parcels they can be like and their
38:14
trucks can have less emissions because
38:16
we're helping to be more efficient with
38:18
with their vehicles delivering those
38:19
parcels
38:20
so in terms of smart cities smart cities
38:23
is a broad definition and we and we're
38:25
leaders in what we call smart mobility
38:27
but it's about reducing congestion in
38:29
cities it's about reducing emissions is
38:31
about improving safety on the roads so
38:33
we're absolutely at the forefront of
38:34
powering all cities becoming smart
38:37
cities with this smart mobility data
38:38
that's super exciting i i have like a
38:40
ton of questions about that actually um
38:43
my so my first big question is all right
38:45
you guys see 20 of all vehicles on the
38:47
road in a specific city for example what
38:50
is the critical mass there is there a
38:51
point where it's like hey when we see
38:53
one in every three cars we can just
38:56
identify all traffic everywhere or is
38:58
that like a bad way to think about it no
39:00
i mean we so we so we've we've built
39:02
some really sophisticated models so we
39:04
so we so we know when we're when in fact
39:06
we're under indexing so
39:08
uh over the whole of the us we see about
39:10
1.2 percent of all vehicles and that
39:13
that as a statistic isn't isn't critical
39:15
mass
39:16
but we've then done uh trend-based
39:18
demographic analysis
39:20
so by virtue of a car being there could
39:23
be more modern it's probably going to be
39:25
more more prime it's less likely to be
39:27
say a a fleet track for example
39:30
so we we over index then um profiles of
39:33
vehicles where we haven't got critical
39:35
mass
39:36
so for us by by understanding by having
39:38
such intimate amount of detail and such
39:40
huge amounts of detail we then machine
39:42
that we've then over we've then indexed
39:43
accordingly so we can now give you
39:46
real-time views of the top 50 cities in
39:48
the us and be statistically relevant in
39:50
all those cities and beyond
39:53
wow that's super exciting and and it
39:55
sounds like so one of the questions i
39:57
was going to ask you was about
39:58
individual pieces of data and where
40:00
they're most valuable but i think the
40:02
question based on what you're saying is
40:04
actually a little different right it's
40:06
you know when you're thinking about
40:08
puddles on the road and sensors catching
40:09
hydroplaning that's useful at the car
40:12
level for things like insurance and
40:13
identifying damage and what happened but
40:16
it's also useful on this macro level of
40:19
in the city identifying where that
40:21
hydroplaning happened and saying hey
40:22
there's a hazard here a hydroplaning
40:25
puddle hazard watch out and send that
40:27
signal to other cars can you explain a
40:29
little bit about like
40:30
what types of data you find most
40:32
valuable in the sense that here's like
40:35
these are the 10 things that we use
40:36
everywhere and all these different kinds
40:38
of scenarios is that the right way to
40:39
think about it yes so we so well one of
40:42
the things we do is we we receive
40:44
nearly a thousand inbound inquiries a
40:46
month from from organizations whether
40:48
it's departments of transport whether
40:50
it's government agencies whether it's
40:52
insurance carriers
40:54
a whole sort of broad array of differing
40:56
industries want to understand about what
40:57
connected vehicle data ev data navy and
41:00
av data could mean for that their
41:01
organization
41:03
and we offer something called uh wejo
41:05
data labs which is a sandbox environment
41:08
and we we we we say look go into this
41:11
environment
41:12
and run queries
41:13
see what you think about the stage and
41:15
it's a trend basis so there's no
41:16
personal information data released but
41:18
it helps us understand about what the
41:21
what the what the recipients of data
41:22
want you know how can we prioritize our
41:24
product roadmap for example
41:26
so i can give you an example of of a
41:28
particular organization where
41:30
um they put four data scientists within
41:32
our within the within ouija data labs
41:34
the sandbox and they run something like
41:36
2000 queries over a four day period
41:39
you know so we got great understandings
41:41
about what how how that data benefit
41:43
that organization
41:45
that particular organization they want
41:47
to understand about whether they could
41:49
optimize in effect real time optimize
41:52
how to reconfigure uh stores based on
41:55
footfall based on based on story traffic
41:57
flow around their store you know should
41:59
they fundamentally change based on the
42:01
profile traffic on a per day on a per
42:04
day basis as to what is at the front of
42:06
the store and you can only do that with
42:08
real time data so fundamentally we start
42:11
with this data lab we let people log in
42:14
we charge them for it um they they then
42:16
run queries we then prioritize
42:18
accordingly as to as as as a broader
42:21
roadmap in terms of value the value of
42:23
data um
42:25
we have we have what we call preferred
42:27
partner status relationships with oems
42:29
where that gives us degrees of
42:30
exclusivity in in defined marketplaces
42:33
unless you or the auto manufacturer has
42:34
a pre-existing relationship in place
42:37
and we've we've determined where the
42:38
high value in data is so we've
42:40
determined that
42:41
traffic management which includes
42:42
mapping is a high value market and you
42:45
may have seen our our our
42:47
our ak
42:49
we talk about the uni economics per
42:51
marketplace per year and by 2025 traffic
42:56
management we talk about over three
42:58
dollars per vehicle per year being the
42:59
value in that traffic space alone remote
43:02
diagnostics rds we talk about that being
43:05
a three dollar mark per vehicle per year
43:07
marketplace
43:08
we then talk about insurance and we've
43:11
broken that into really great detail one
43:13
thing that we will be starting to report
43:15
in our 10 qs and our 10k
43:17
is is the unit economics per vehicle per
43:19
year per territory per marketplace so
43:22
complete transparency
43:24
so and it helps you understand you know
43:26
the people ask me what what's the
43:27
biggest risk for ouija and obviously
43:30
regulatory is always a risk but actually
43:32
our biggest risk is opex
43:33
it's are we getting the timing right for
43:35
the right marketplace in the right
43:37
territory for the right group of oems um
43:40
we have demand already to build
43:41
insurance products in japan is it right
43:44
to prioritize the rollout in japan
43:46
compared to say europe for insurance
43:48
well we've got the commercial
43:49
justification to do that but there's
43:51
still we're still front loaded in terms
43:52
of our opex i'll spend before before we
43:55
see it before we see a payback so it's
43:56
important we get that right and we
43:58
maximize ultimately the unique economics
44:00
per vehicle per year per territory
44:02
no that makes a lot of sense and you
44:03
know maximize that while minimizing your
44:05
risk so it sounds like the big lesson to
44:07
investors is what we joe is aggressively
44:10
tackling now is understanding at the
44:12
unit economics level what data is
44:14
driving the most value across a wide
44:16
variety of verticals for example looking
44:19
at hydroplaning data not just at the
44:20
micro one vehicle level but at this
44:23
macro you know smart city hazard level
44:25
and identifying which sensors contribute
44:27
to that
44:28
through you know the journey of the
44:30
entire data and focusing aggressively on
44:32
user using and optimizing those use
44:35
cases that's where the value is and
44:36
that's how that's how we try and
44:38
differentiate ourselves is by not saying
44:40
we can do everything today there's loads
44:41
of value in it today it's actually do
44:43
you know this is a long journey
44:45
but but over time as as we as we build
44:48
up that value that that value per
44:49
vehicle you'll see our unique economics
44:51
slowly scaling up which is a great
44:53
recurring revenue model and it's a great
44:55
compounding sort of value to the ibm's
44:57
where the typical vehicle has got a
44:58
lifetime of seven to nine years so even
45:00
though the stock market isn't favoring
45:02
high-tech growth companies right now we
45:04
joe is partnering with big players like
45:05
microsoft and palantir to solve some big
45:08
problems for over a dozen major car
45:10
companies today from ford and gm to
45:12
lucid and even bentley in my opinion
45:15
their path to growth is clear as
45:17
vehicles get more and more sophisticated
45:19
each one will have more data to provide
45:21
we joe that compounds with more and more
45:23
of them getting connected to wijo each
45:25
and every day add in smart
45:27
infrastructure micro mobility solutions
45:29
and digital twins of everything that we
45:31
joke connects to and you have a very
45:33
powerful end-to-end platform for
45:35
connected vehicle data all of that data
45:37
will unlock new use cases and ultimately
45:39
new markets for wijo to enter and use
45:42
their massive amounts of data for good
45:44
that's a future i'm really excited to
45:46
keep digging into so on behalf of ticker
45:49
symbol you a big thank you to ouijo's
45:51
founder and ceo richard barlow for doing
45:53
another exclusive interview as well as
45:55
the rest of the ouijo team for making
45:57
this happen during such an exciting time
45:59
in the stock market and as for you stay
46:01
long stay strong and thanks for watching
46:03
until next time this is ticker symbol
46:05
you my name is alex reminding you that
46:08
the best investment you can make is in
46:10
you