Sign up to save your podcasts
Or
Share Building the AR OS for Enterprise, with RE’FLEKT’s Dirk Schart
Share to email
Share to Facebook
Share to X
Share to email
Share to Facebook
Share to X
Or
Building the AR OS for Enterprise, with RE’FLEKT’s Dirk Schart
Most businesses
have the information and infrastructures they need to be more
efficient and competitive — it’s just a matter of having it all at
their fingertips. RE’FLEKT is working at making that process easier
by creating a modular, scalable, open-source operating system for
businesses to build their own in-house AR applications on top of.
President and CMO Dirk Schart drops in to explain how.
Alan: Welcome to the XR for
Business Podcast with your host, Alan Smithson. Today’s guest is Dirk
Schart RE’FLEKT President and CMO. Dirk Schart is a marketing and
growth driven tech executive, whose current mission is to build the
operating system for enterprise augmented reality. Dirk leads the US
operations for RE’FLEKT. Dirk is one of the first employees of
RE’FLEKT and is a funded company by global companies such as Bosch
and BASF, and today plays a leading role in enterprise AR.
Previously, Dirk built the Digital Innovation Lab for Hyperloop
Transportation Technologies, for which he continues to work as an
advisor. Within four months, he built a team and developed the first
MVP, and he also directs the VR and AR team for Hyperloop
Transportation Technologies. Dirk helps technology startups such as
VisualX, VYON, and RiseStep. He’s the author of two augmented reality
books, albeit they’re both in German, but you can get them. And he
also contributed to Metaverse, the book by Forbes writer Charlie
Fink. He founded the first AR blog in Germany, “WE ARE AR”
and has been interviewed and quoted by leading media such as Tech
Crunch, Venture Beat and Huffington Post. For more information on
RE’FLEKT, you can visit re-flekt.com.
Dirk, welcome to the show.
Dirk: Hey, Alan. Thank you very
much for having me.
Alan: My pleasure, my friend.
The last time I saw you was at Augmented World Expo, and I think we
were taking silly photos.
Dirk: That’s true. But was fun.
[laughs]
Alan: It’s always fun. And it’s
becoming like a family, this whole augmented reality family and
everybody’s working together. And it’s been wonderful. And the work
that you guys are doing at RE’FLEKT is world class. And I really want
to dive into the benefits of this technology and how you guys are
using it. Let’s talk about that.
Dirk: Absolutely. So you already
mentioned it in the intro; what we’re doing is we’re building the
operating system for enterprise AR. So what does that mean? That’s
for us the foundation of how enterprises will use AR in the future.
Let me go back quickly and talk a little bit about how we started.
That makes it easier for the listeners. Back in 2012, it was really,
really difficult to create any kind of augmented reality
applications. So it was completely different to what we have today.
And we started to develop individually programmed applications. There
were no platforms. There was nothing at the time. And one of the
first things we did is, we build an application for Range Rover. And
the reason for that was that they came and said, look, we have a very
complex repair. It’s about a few return line of a Range Rover car.
And the mechanics, they always have problems doing that because it’s
complex. They don’t do that every day. Documentation doesn’t seem to
be perfect for it for that use case. They try PDF, they tried videos,
but result was the same, they couldn’t fix the problem. So we built
an application.
We showed it to them and helped the
mechanic, because it guided them step by step through that repair and
showed them exactly at the point where they had the problems. “Wait.
Now focus on it. Have a look. Double check before you do the next
step.” And that at the end helped them to perform the task much
better and reduce their error rates. And then, of course they asked
about, “okay, so how can we do that for all our car models?”
And at the time — that was at 2013, 2014 — there was no scalable
way to do that. And that was for us the starting point to build our
platform, which is called RE’FLEKT 1, which allows enterprises to use
your existing content.
We’re talking about everything
enterprise have already, because it’s existing. They have it in your
documentation, in a technical documentation for owner’s manual to
repair manuals, maintenance manuals, all that stuff. It is somewhere
in the cloud or on a server, but not used for AR. So we enabled that
the enterprises to reuse that with a platform. And like that, we
started and we rolled that out. We added remote support and things
like that. And over the years we realized, okay, now that that’s
working even if there are a lot of challenges and you know that,
we’re not there to say it like that. In a way they say, “well AR
is kicking off in a way that everybody’s using it with the CRM.”
But the next step for us, and now coming back to the initial point
about “what does it mean to build the operating system for
enterprise AR?” is, to have a platform is one thing. But there
are many vertical specific requirements today. Whether you go into
medical or pharma, where you have regulation, or you going to
automotive or aerospace. The scenarios are completely different, and
the requirements are different. And we as an AR startup — and that’s
the same for other AR startups — we cannot solve all of these
problems. We are technology companies. So this is why we said we take
our platform, we make it modular, we add SDK and API to make it open,
and building that operating system so that partners, system
integrators, solution providers, but also enterprises on their own
can build solutions they need. And this is exactly what companies ask
us for: to have a solution where they can customize their stuff,
where they can integrate into their solutions. The funny thing is we
always think as AR company and you mentioned it, it’s a community and
it’s cool. We are all so cool in AR, but in enterprise business,
they’re not going to change their infrastructure and their
architecture and the systems because of AR. It has to happen the
other way around. And that’s why we’re doing what we’re doing.
Alan: What do you mean by that?
“They’re not going to change their infrastructures is off
because of AR. It’s the other way around?”
Dirk: What we tried is to offer
our solutions and then go into the enterprise and say, “Well,
look, here it is. You have an AR solution and now you need to
integrate into that. You need to provide your data and all of that.”
But enterprises have existing infrastructures and architectures. They
have a CRM, they have an authoring system, they have a PLM system,
they have a CAD tool. They have all of that. So AR is just one little
thing, and AR has to integrate into those kind of systems, not the
other way around. No company is going to change a lot of their
infrastructure or their CRM or their CAD tools because of AR, that’s
not going to happen. And that’s exactly why we deliver it as an
operating system.
And now we allow and say, okay, for
instance, we connect to Siemens PLM Team Center. So that means a user
who already works in something like Team Center does not need to
leave that environment. Our platform kind of runs in the background.
It is used for publishing, for helping with the technical stuff,
making that easier, like building a tracking configuration. All of
those things are integrated into our platform, but the author can be
at the used environment which he or she already knows. So it’s not
about learning all of that new. It’s just to have kind of an AR layer
on top of it. And that makes it much easier. And that’s — for me and
that’s for us — the key to make AR scalable and also to have finally
acceptance on a higher level. Because the technology is available,
it’s not a problem at all. We can realize all of those things the
enterprises need. It’s much more about the acceptance among IT people
and also among the frontline workers.
Alan: Interesting. Yeah,
somebody mentioned on one of the other podcasts that we don’t have a
technology problem anymore. We have an adoption problem.
Dirk: Right. That’s how it is. I
fully agree on that. Well, look, as I said, we can realize it. I
mean, give you an example. We had one little gap in the publication
process. You could do everything. You could upload your data, you
could add your AR content. And then at the end you could publish. The
only thing you could not do was to build your own tracking
configurations. So it means you need to tell your AR system and your
camera and your algorithm when it sees a machine. What does the
system have to place on top of the machine? What is said to be
overlaid here? And that’s called a tracking configuration. That had
to be done by an engineer or by a software developer. We closed that
gap; we created a tool. And with that tool, everyone can do that,
without any kind of coding skills. You just use your existing data.
You compare it where now expressing a very simple way and that’s it.
So it closes the gap.
That supports what you just said. We
don’t have a technology problem. We can fix all of that. Now also
with machine learning on top of it, which improves the tracking and
the usability of the whole system. But don’t forget that, especially
when we talk about the frontline workers, the actual users, then we
have to talk about change. And you can just go and say, “Well,
look, you don’t use your hammer anymore, now we give you an iPad and
that’s it.” That does not happen from today to tomorrow. We have
a new report we worked on, why so many POCs or trials get stuck in
that early process and do not proceed to a deployment, to a rollout,
what we all want to have in the AR community. And it’s about change,
it has nothing to do with the technology, maybe with the usability.
But at the end it’s change, and you need to tell the people why they
should use it, what is the advantage, and how can they use it?
Alan: All right. So let’s unpack
that. Why should customers use this? How do they use it? And then
what are they going to get out of it?
Dirk: Talking about the why,
they’re the obvious things, which we repeat over and over. Talking
about enterprise use cases, you can reduce your downtime, you can
improve your first time fix rate, reduce your error rates. There’s so
much more than only that. I’m thinking about that you can really
improve your customer experience. You add a lot of value across the
product lifecycle, things you could not do before. And I said when I
explained that example, when we did it for Range Rover, that that
showed it clearly. Take it from the consumer side: you buy a new
product — let’s take a coffee machine. So you unpack that. What do
you want to do? You just want to use it. You don’t want to flip
through pages and read how that works. And it’s an awkward
experience. It takes you forever until you get that coffee, finally.
Alan: Don’t mess with people’s
coffee.
Dirk: Yeah. [laughs] My coffee
machine did not work in the morning. I’m lost without my coffee! But
nobody wants to read the printed manuals anymore. If you find them,
then they’re outdated. It takes a lot of time to understand what’s
written there. You don’t want to have that anymore. And let me add
that to the why. I think it has a lot to do also with what we know
from consumer side. Even that many people say, “Oh, well, you
know, but it’s enterprise, and enterprise is okay to be boring.”
No, it’s not. We are used to Netflix. We’re used to iPad. That kind
of experience. And then we go to work, and then we have to use the
old tools which do not provide anymore what we need.
Alan: It’s crazy, right? We have
these beautiful UX/UI for everything in the consumer world. And then
you go into these enterprises, and it looks like Windows 91.
Dirk: [laughs] Exactly! And
that’s the problem. People don’t want to have that anymore. You come
from home and you go to work and you have to go into another world.
And that’s not acceptable anymore. And that’s what we see, that
transformation happens. And also we think about the next generations.
If I see my son — and he’s eight years old — and I showed him the
computer and I gave him a computer mouse and he looked at me and
said, “Dad, what’s that?” And I said, “that’s a
computer mouse, for your cursor here.” And he just left. And I
said, “What?” He said, “No, I don’t want to have that.
I will use my iPad.” That’s a change in behavior. And that’s why
I say we need to bring the consumer grade experience to the
enterprise sector. And AR helps us a lot to do that. Now coming to
the how, well, that’s a bigger part. I mean, we could do several
episodes here about how. But let me quickly given all of you about
that. I think it’s a lot about how you can get started. In the past–
the last years, many enterprises kind of wanted to get started with a
huge thing. Everybody was thinking about that rollout. How do you
decide that if you were a manager in a large enterprise, are you
going to decide and say, “I will invest 500k into a new
solution, where the ROI is not proven or I don’t know what’s going to
happen exactly. I see there is a potential, but I don’t know exactly
how it’s going to work. And I have to report that.” Nobody’s
going to make that decision. And that’s, I think, one of the problems
we had in the past.
Alan: I always tell the story
how when we started meeting with customers, when we started, they’d
go, “Great! This is amazing technology, I love it, it’s really
mind-blowing! Who else is doing it? What are the ROIs and how much do
they cost?” And we would say “Nobody else; we have no idea;
and a lot. Still want to buy it?”
Dirk: [laughs] Yes, that’s
reality. Not augmented; only reality. That’s exactly how it is,
right? And those are the questions that we have every day. It’s not
enough anymore to have some interesting logos on your slides and say,
I work with ABC. No, you have to prove it. How many users do you
have? How did they do the rollout, as you said it? What is important
is you need to do it in steps. You need to start with that with a
trial. Go do a POC–
Alan: Ok, so what are the steps?
I’m a new customer. I own a factory. You guys come in. What’s step
number one?
Dirk: So step number one is, for
us, we start with a little package and do a usually a scoping
workshop, where really to check what are the requirements? Is there a
company who just wants to test something? Is there a company who’s
already thinking about a bigger solution? And then specifically, what
is the use case? I really want to see a use case when we get started.
It does not need to be any complex thing, but just one thing. And
then in the workshop, we define “Okay, that’s the use case.
That’s the way we do it today. And then we have KPIs. And that’s the
way we’re going to do it with AR.” And then we can measure.
That’s the key. So and that gives the enterprise a small package
where they can define what they do together and with the front line
workers that’s important. It needs to go bottom up, not only defined
by the management, because that will be a failure. And then you have
that scoping and then you have the measurement. You can say, “Okay,
this is how we did it before, and this is how you do it with AR. And
now we can see that’s successful. Does it make sense in that use case
[garbled]
?” So once you have that, you go to the next step and
you go into a POC where you can really test that internally, within
several machines or several products and really start to scale it.
Not only in a test environment, but in a real world scenario, which
makes it different or even involve your customers and get the
feedback there. And then that’s the second step. And then to come to
the third step where you finally roll it out. Take all the experience
you have from step one and step two, take that into consideration.
And then finally, you’re able to roll it out. Because between 1 and 3
dictate a lot of different steps.
Step 1, it’s more about figuring out
how it works, maybe already involving the IT a little bit and say,
“Okay, so what kind of connections do we have here? How do you
install all of that?” And step 2 is then completely focused on
the business case itself. “Who’s going to use that? How is that
compared to the current solution? What is the customer or the user
feedback?” And step 3. It’s a tough part to roll it out. It’s
about security. It’s about integration. It’s about all of those
things. And that’s a completely different story. I mean, it takes you
a month to really get the security done and get approved as a vendor
with a large enterprise to fulfill all the requirements.
Alan: So this scoping workshop,
POC rollout, is this a 12-month rolling timeline? Or longer, or
shorter?
Dirk: It depends a little bit on
the enterprise side. There are some there, they’re faster. But there
is some that need more time. So bringing that into a timeline that
the scoping workshop and the results out of that. That’s something we
do. Let me say maximum 60 days. That’s the timeframe. What we also do
is, we do the scoping but we also give them the product to work on
that, to get a feeling. What can they do? How can they create content
on their own or with our support? So we really train them, we have
our own education program where we train them. So our promise is up
to 60 days. A company knows how to create an application on their own
and they create the first application on their own. And then, for the
POC, you can add another 90 days, because that has to be prepared. So
60 or 90 days for the first two parts. And then rollout, I would
definitely plan a six month time period. That’s what you definitely
need to prepare all of that. It’s a lot of communication and a lot of
things to check internally. Month in, you can finally roll it out.
And even then, you probably do it in a few countries or with a few
users before you do that globally. It’s a big thing, to do that you
have a lot of different requirements. Might think about things like
GDPR we have in Europe, which you do not have. And the same in the
States or in Asia. Think about Asia and China. You cannot just go
there and host a server. It’s just not possible. We tried that and
luckily we have a partner like Bosch, our investor and partner,
Bosch, and they helped us with lots of stuff there, because you
cannot just go there and say, “OK, here we are. No, let’s get
started.”.
Alan: I guess not.
Dirk: No, it’s not. It’s not
that easy. It’s possible. And the market is growing fast.
Alan: What are the KPIs? What
are the measurements that we’re going to measure? What are some of
the metrics that that you would test against them? And is there a
cost to phase 1? What are the costs associated with phase 1, phase 2,
phase 3?
Dirk: Yeah. With phase 1, it’s
usually a wrench between, let me say, $10k and $20k. POC and rollout
is a little bit harder to say. POC depends on what exactly is the
scope of it. Let’s say there is between 20 and 50. But in rollout,
it’s hard to put in numbers, because that’s very specific and depends
on what kind of solution you want to have. But of course, then it
goes into licensing and subscription. That’s tough, but it’s also a
50-50 plus then, to have that in numbers. So KPIs you asked for. So
that’s something that– the interesting thing is the enterprises that
they know their KPIs, they know exactly what they have and what they
need. The typical things, what you have is error rates, first time
fix rates, down times, when you think about the manufacturing sector
especially. That’s a very clear KPI. But of course, there are many
more if you look into the training sector. How do you measure
knowledge obtainment? How can you measure knowledge transfer? When we
think about the skill gap and the workers gap in the manufacturing
sector. And they have the big issue that they need to transfer the
knowledge from the people getting retired to the young guys.
Sometimes you have really hard numbers. Like what I said before, they
reduce time to downtime or the error rate. And sometimes it’s a
little bit more difficult to say, “OK, how can we measure things
like knowledge transfer?” and we have to define that together
with the customer. It’s very important to find a base, to compare
that. We had a customer and they did it a completely different way.
We asked, “So how did you measure the success?” and the guy
who led the project, he said, “Well, it was simple. I asked my
team what they think, and the team said we’re super happy, and it’s
better to work than before.” And I said, “Okay, I
understand. But what are the metrics behind that?” And he said,
“No, there are no other metrics. It’s just the team is happy and
works better than before.” I said, “OK, that’s an
interesting way to define a KPI.”
Alan: [laughs]
Dirk: Well, this shows you we’re
still early stages, right? We’re building KPI frameworks for our
customers. But it’s not that you have one framework which you can use
for all kind of customers. It’s just [garbled] word right now. I had
a meeting last week with a large enterprise in Florida and for them,
other things were super important. For them, it was important how can
we automate the content creation? How can we customize the UI, that
it fits to what we have already? We want to integrate that. We don’t
want to have that our service technicians have to get used to a new
UI or something new. So that was for them much more important than
other matters. They said, “Well look, that’s what’s for us now
super important, because that helps us to integrate the system and to
show the employees and the workforce, well look, that’s a good tool,
use it. And then we will look for the other KPIs once they’re using
that.” And I like that approach because as I said, it’s about
change. So the first step, you have to make sure that people want to
use it. What do we have right now in 70 percent of the cases, we
create a tool which is very helpful, but maybe workers or technicians
use it once a week, but they don’t have that frequency that they use
it every day. So it does not have the value it’s supposed to have. It
really depends on what is your goal. And then you can define your
KPIs.
Alan: Now, you’ve run a number
of these in this protocol. So this 1-, 2-, 3-step process — scoping
workshop, POC, and then roll out. How many of your customers are in
the rollout phase now? Or maybe already rolled out?
Dirk: Several are in the rollout
phase. The interesting thing is — and that’s also the sad thing for
us enterprise AR guys — is that were not allowed to talk about all
that stuff.
Alan: [laughs] I figured that.
Dirk: That that’s one of the
most difficult things. But we’ve already rolled out. So like the
biggest rollout we had was an application, which is called Daimler
Rescue Assist. And I think it’s probably one of the biggest rollouts
in AR history, because it is available in 27 languages and for all
Mercedes cars since 1990.
Just to give the background quickly,
why it is so big: it’s an application for first responders. And when
they come to an accident, they have to figure out where are the
critical parts like battery wires, fuel tank, things like that. And
they have to make a decision, in seconds, where they cut into the car
if there’s someone inside. And before, they had to do that with PDF
instructions, and you can imagine that’s quite tough if you have to
compare that. You have deformation in the car. Then you have to cut
into that car. That’s not really a good decision-making basis. That’s
the reason why we have that application. And Mercedes and Daimler,
they wrote that out globally, together with our partner Bosch. So
that’s one of the big examples. And that shows also it’s possible to
do that. And the others, we have currently between five to ten
rollouts in preparation, communicate more about that soon. There are
different companies from the chemical sector, but also from the
energy sector. And what you can see is, it’s completely different.
There is not one rollout. I mean, you have some things that always
happen, it’s like security as the major topic–
Alan: What are some of the
things that are always consistent? From what I’m hearing, you’ve got
one major rollout, a couple more coming. So, we really still are in
the earlier stages of this technology. And again, it’s an adoption
issue, not really a technology issue. But that is only recent that
the technology has caught up to our visions.
Dirk: Yeah, absolutely. I mean,
I think it’s a lot about once we have seen the first five, 10, 15 —
that will be the trigger. I think many enterprises are waiting,
they’re ready to go, but they’re waiting to see others. And then they
will quickly turn into the–
Alan: Fast follow.
Dirk: Yeah, exactly.
Alan: I live in Canada; we’re a
country of fast follow. We watch what America does, and we wait for
them to make their mistakes, and then once it’s proven, we jump on.
Dirk: That’s like the Apple
strategy, right? [laughs]
Alan: Exactly. Never be the
first; always be the best.
Dirk: Exactly. But that’s what
we see. They want to seed, they invest, but they don’t want to make
the decision about the really big deal. And then on the other hand,
also, it is a big deal to do a rollout like that. And if you have
like several thousand of users involved there, a lot of questions you
have to answer. And that starts. You have to involve like branding,
who is going to host your app? Who’s going to do that? What our
security requirements and all of those things. Where are we going to
start? Which country is the first one? What kind of devices do we
have? And don’t forget about by large enterprises.
We often have old operating systems,
whether it is about Windows or even tablets and stuff like that. So
you cannot really compare it. There are some basic patterns. Security
is one. I would say less like 70 percent of the securities that
that’s basically the same large enterprises. Then there are some
specific rules. But basically that’s the same. So what you have done
that. And you know how to handle that and you have the right tools
for it, then that becomes easier. But the other things that’s really
different because organizations are different.
Alan: So what are some of the
considerations? Let’s just run through a list and we’ll just compile
a big list and we’ll put it in the show notes for people so they can
at least have a list of considerations. Right now I’ve got security,
branding, hosting location devices, device management, integration.
Dirk: Right. Countries. That’s
something we have to check because there are different devices we
already mentioned. Those are the well precise and the league terms in
terms and conditions. So that part is very, very important. How do
you do that? And then it depends also on is that an internal pronoun
or John Roland where he goes to the customers? That’s also completely
different. That’s what we what we have right now, where the customers
are invoked. Then you need to think about, okay, how do we provide
access to customers? What kind of systems do they use? So it becomes
kind of a narrow level of complexity on top of that. But you learn a
lot in different units, have different requirements. And then I think
one of the biggest parts when you when you get started with that is
to find the right people for it. You really need the right people in
these large enterprises who know how it works. That’s for me, always
a major point. We come from our AR world and we think we know
everything and then we come to an enterprise, and the first thing we
have to do is we have to listen. Because we have to learn about, how
does that enterprise work? What are the structures? How is the
organization organized, and what were the specific things they can
tell you? There are so many different things that they have to look
for. And then it’s about customization. If you can, define steps for
the rollout. But that does not mean that it really fits for all
enterprises. You need to have an individual plan for an enterprise to
do a rollout, that it fits. Otherwise, it’s not going to work.
Alan: All right. So what are
some of the challenges that you guys have seen that you’ve had to
overcome, over the time that now, is just part of your process? What
are some of those challenges that people getting into this [can
expect], that seem to be general challenges?
Dirk: So I think one of the
biggest challenges is what I mentioned earlier: the way we get
started. Instead of saying, “well, look, here is a license,
here’s the product. You get a training: now get started. Create
applications and roll them out.” That’s what we thought we can
do. Just more or less, you sell a CRM and enterprises install it, and
they get started. But we’re not selling a CRM. We’re far away from
selling CRM. When you buy a CRM, it’s more about, “which one do
I need,” but not whether you need one. With AR, it’s about, “doI
need AR: yes or no?” It’s not about which kind of solution, the
first step. And most of the enterprise, they even don’t think about
it. And most of the enterprises I see, they even don’t know about the
different solutions or vendors. That was one of the biggest
challenges for us to learn and understand. What is the right model to
get an enterprise started? And large enterprises, mid-market,
industrial market — it’s just the same. There’s no difference from
that point of view. So that’s why we came up with the different
stages, right? Scoping workshop, POC, and then rollout. That helped
us a lot to learn together with the enterprises, and take the
learnings, improve it, and then do the next step. Enterprises are not
ready to buy a license right away and get started with it. There are
[some]
, of course, but that’s not the typical way to do that.
Alan: So, you’re a platform.
Dirk: Right.
Alan: You’re also a content
provider, and you’re also a service provider. You’re kind of having
to do everything, is that what I’m hearing?
Dirk: Yes.
Alan: That will change as this
technology becomes more adopted, and people can start to realize the
benefits of it. But I think where we are right now in this
technology, from what I’m hearing from you, is that we’re still so
early that we still need to handhold every customer through the
process of deciding that, “this is something we want to do,”
and then trying it. “Yep, it does do what we say it’s going to
do,” and then, “okay, it’s time to roll it out.” But
the companies are still ill-equipped to do this themselves. Selling
just the CRM system is not possible just yet, but it will be in the
future, as more and more case studies come out and people are able to
say, “okay, I want that,” right?
Dirk: Exactly. That’s what you
describe is reality today. The interesting thing is that applies for
all our startups. If you have investors, and investors — of course
— look for recurring revenue. So, you shouldn’t say, “okay, we
will sell licenses and nothing else.” But in these — and that’s
not only for AR; I see exactly the same in IT, in block chain, in AI
— we are not at a point I can just sell the licenses and the
customer takes it and does whatever. We need to have the education
and the service part. There are many, many enterprises. They ask me,
“what kind of person is that, who would create the content? What
kind of skill set do they have need to have? Whom do I need to hire?”
That was, for us, the trigger, and we started now with our own
education program, where you can get your own modules, and companies
can book that, and they can train and educate their own people to do
that, [and ensure] that they have also the same level of education
and same level of knowledge. We need the services until, as you said,
we reach a point that everyone knows. But even look at the big
systems. If you look at SAP, or even a CRM — take salesforce,
something like that — you need a lot of customization. You need a
lot of service to implement that in the right way. And AR is no
different.
Alan: It’s interesting you
mentioned that, because this is one of the reasons we started XR
Ignite, of which you’re a mentor, as well. XR Ignite was started as a
community hub to foster the relationship between startup studios and
developers, and corporate clients. Because what we realized — same
thing you just mentioned — is that the content creation is a
separate thing from the platforms, and venture capital invests
typically only in products and platforms. But what they’re missing is
this sweet spot of content creation. And one of the things that we
realize is that independent developers and small studios, they’re
going to start to get acquired, because companies — exactly what you
said earlier — they’re going to say, “who do I need on my team
to build this content? Oh, I need 10 people with Unity skills. Well,
I don’t have time to hire 10 people. Why don’t I just acquire this
little studio here that can do what we need, or work with them for
six months? Test them and then acquire them.” I feel that
there’s going to be a lot of these small micro-acquisitions of $5-,
$10-, $20-million acquisitions, where teams are just basically
acquiring talent.
Dirk: Absolutely. And you can
see that already, that companies are thinking about that. How can
they solve that? Do they have own teams mostly? No, they don’t have.
They have like one or two–
Alan: No, and it is going to be
hard, because right now, there is a shortage of this talent. And if
you project out two years, when all of the companies all of a sudden
wake up, the talent pool is not going to be much bigger, because
schools and universities and colleges are not teaching it yet. And I
think we’re going to find this issue right across the board with
exponential technologies, that our current school systems are
ill-equipped to retrain people. How do we train people fast enough to
become very, very proficient in creating this technology, so that
companies can start to either acquire studios or startups and really
build them in, or hire them?
Dirk: Absolutely. And that’s
what we need. And it’s different often, right? I mean, content
creation is something… especially when it comes, now, to the two
big things we do: repair and maintenance in areas, in training
environments, and in the manufacturing or medical sector. You take
your CAD — people know how to do that part — and then you add your
AR layer on top of that. But then it already starts, and is okay, how
can I do that? But I think is there is a shift from 2D people to 3D
people today. Content creators, developers. I used to build 2D
interfaces. What we have in our mobile phones; tablets. But now,
handling that in the 3D space and make that a great user experience
and guide the people and help them to use it? That’s a completely
different story. And we don’t have that. There are almost no people
— think about UX designers — being able to think in a 3D space in
that way, especially in the enterprise sector. There may be in the
consumer area, but in the enterprise sector, there is a lack of
people.
Alan: I agree, and I would
punctuate that with the fact that ARKit — which is Apple’s framework
for developing AR — and ARCore — which is Google’s — majority of
the really interesting, robust content that’s coming out of ARKit and
ARCore is consumer-driven. It’s games, it’s experiences. Even if it’s
a business application, it’s usually some sort of marketing. These
applications — while also very, very valid — designing enterprise
AR is not as sexy as designing a video game. It’s more lucrative, but
less sexy. So you’ve kind of got these game developers over here that
are making amazing games in three dimensions. But there’s a skills
gap in between the enterprise and the gaming world.
Dirk: Yeah, you need to know how
enterprises work. You need to see how one of these frontline workers
is using the tools. And that brings me back to how we started with
AR. I remember some of the apps, where we showed a car mechanic how
to change a tire. I mean, that’s not going to work. We need to see,
where does AR make sense? Where does it provide value? And that’s
something that you have to combine that with a good user experience,
because it has to be interesting. It has to be also fun, even though
it sounds weird for an enterprise sector. But it is like that, coming
back to what we what we discussed earlier; you need to have that
Netflix and Apple experience to have a good work result. And that’s
completely different to what you do for games. You can learn from
games, right? Not to underestimate that. You can learn a lot from
games: how to keep people interested, how to keep people at a level,
how to provide information and transfer information. There are so
many different things. See a firefighter: how they hold the tablet
and use it with their thumbs. Maybe they have gloves even. Then go to
a car mechanic. This is a completely different story, and that’s what
you have to understand, too.
Alan: Oh, man, there’s some so
much here. So let me ask you, on a personal note: what is one problem
in the world that you want to see solved using XR technologies?
Dirk: Well, that’s a great
question. There are a lot of things I see, but I think for me, even
that it sounds a little bit global now. What I want to have is I want
to enhance my work environment. I don’t want to be tied to little 2D
screen. That’s where I see the biggest impact of VR. Give me my 3D
work space, where I can define my environment and I can take it with
me. And I always have my office with me wherever I am without having
a little 2D screen.
Alan: And then your office does
not need to be limited to a desk with four walls anymore. You could
create an an an environment… like, I like working on the beach in
Bali. So, every time I put my headset on, I’m in Bali on a beach,
working on a beautiful desk with 100-inch screens around me. The one
thing that we don’t have yet is the ability to have a keyboard and
type in the area, which is crazy. One of those startups that applied
for XR Ignite, actually, is working on that. They’re working on a
keyboard that allows you to see your fing– well, not your fingers,
but see your keys in VR. And it allows you to type in a more
comfortable way in VR.
Dirk: That’s definitely
something we need. And the things I’ve tested so far were not really
usable in the way I want to have it. So right now, I think we only
can enhance our current desktop office, but we really cannot replace
it. But what you mentioned is exactly right. I mean, you see that
often: the keyboard is really a key element in the way we work.
Absolutely. Yes.
Alan: Yeah. And actually,
Logitech introduced their their three dimensional pen recently. I got
a chance to try it at Augmented World Expo. And then there’s another
company called Massless. And basically what they’ve created is an
input device, looks like a pen, but it’s tracked in 3D space. So now
you can draw in three dimensions; something like Z space, which is a
computer that allows you put on really inexpensive glasses and see
three-dimensional things come out of the screen. They’ve also got
this kind of stylus effect, where you can use a pen to grab things
and bring them into 3D space. So I think the user input devices — be
it a keyboard, or voice, or gaze even (looking at something), or a
pen — we’re only just starting to figure these things out.
Dirk: Yeah, we’re early stages.
On one side, we want to reinvent the whole thing, which makes sense
because we need it. It’s a different environment. On the other side,
we also need some elements which we’re used to. Right? I mean, that’s
always funny. Many people also said, “so you tested a lot of
variables; a lot of AR glasses. So, why do we always have the typical
rectangular screens in that 3D space?” I said, “well, look
it’s just the beginning. But also keep in mind, if you learn
something new — if you have to get used to something — then maybe
so it’s good to have some things you are familiar with, right? And a
rectangle screen, with an X at the top right corner that lets us
know, okay. I know how to deal with that. And that’s also a good
feeling.” So it’s a transition phase somehow, where we have to
learn. And it’s interesting that you mentioned the stylus. Yes — I
played a lot around with the Apple pencil. Everything is digital
today. And I came back to that and I didn’t want to use paper,
because I don’t want to have another notebook in my backpack. So I
use that. And I have to say, it’s really a good way to work. I can
imagine that it helps us to work in space.
View all episodes
By Alan Smithson from MetaVRse
4.5
1212 ratings
Most businesses
have the information and infrastructures they need to be more
efficient and competitive — it’s just a matter of having it all at
their fingertips. RE’FLEKT is working at making that process easier
by creating a modular, scalable, open-source operating system for
businesses to build their own in-house AR applications on top of.
President and CMO Dirk Schart drops in to explain how.
Alan: Welcome to the XR for
Business Podcast with your host, Alan Smithson. Today’s guest is Dirk
Schart RE’FLEKT President and CMO. Dirk Schart is a marketing and
growth driven tech executive, whose current mission is to build the
operating system for enterprise augmented reality. Dirk leads the US
operations for RE’FLEKT. Dirk is one of the first employees of
RE’FLEKT and is a funded company by global companies such as Bosch
and BASF, and today plays a leading role in enterprise AR.
Previously, Dirk built the Digital Innovation Lab for Hyperloop
Transportation Technologies, for which he continues to work as an
advisor. Within four months, he built a team and developed the first
MVP, and he also directs the VR and AR team for Hyperloop
Transportation Technologies. Dirk helps technology startups such as
VisualX, VYON, and RiseStep. He’s the author of two augmented reality
books, albeit they’re both in German, but you can get them. And he
also contributed to Metaverse, the book by Forbes writer Charlie
Fink. He founded the first AR blog in Germany, “WE ARE AR”
and has been interviewed and quoted by leading media such as Tech
Crunch, Venture Beat and Huffington Post. For more information on
RE’FLEKT, you can visit re-flekt.com.
Dirk, welcome to the show.
Dirk: Hey, Alan. Thank you very
much for having me.
Alan: My pleasure, my friend.
The last time I saw you was at Augmented World Expo, and I think we
were taking silly photos.
Dirk: That’s true. But was fun.
[laughs]
Alan: It’s always fun. And it’s
becoming like a family, this whole augmented reality family and
everybody’s working together. And it’s been wonderful. And the work
that you guys are doing at RE’FLEKT is world class. And I really want
to dive into the benefits of this technology and how you guys are
using it. Let’s talk about that.
Dirk: Absolutely. So you already
mentioned it in the intro; what we’re doing is we’re building the
operating system for enterprise AR. So what does that mean? That’s
for us the foundation of how enterprises will use AR in the future.
Let me go back quickly and talk a little bit about how we started.
That makes it easier for the listeners. Back in 2012, it was really,
really difficult to create any kind of augmented reality
applications. So it was completely different to what we have today.
And we started to develop individually programmed applications. There
were no platforms. There was nothing at the time. And one of the
first things we did is, we build an application for Range Rover. And
the reason for that was that they came and said, look, we have a very
complex repair. It’s about a few return line of a Range Rover car.
And the mechanics, they always have problems doing that because it’s
complex. They don’t do that every day. Documentation doesn’t seem to
be perfect for it for that use case. They try PDF, they tried videos,
but result was the same, they couldn’t fix the problem. So we built
an application.
We showed it to them and helped the
mechanic, because it guided them step by step through that repair and
showed them exactly at the point where they had the problems. “Wait.
Now focus on it. Have a look. Double check before you do the next
step.” And that at the end helped them to perform the task much
better and reduce their error rates. And then, of course they asked
about, “okay, so how can we do that for all our car models?”
And at the time — that was at 2013, 2014 — there was no scalable
way to do that. And that was for us the starting point to build our
platform, which is called RE’FLEKT 1, which allows enterprises to use
your existing content.
We’re talking about everything
enterprise have already, because it’s existing. They have it in your
documentation, in a technical documentation for owner’s manual to
repair manuals, maintenance manuals, all that stuff. It is somewhere
in the cloud or on a server, but not used for AR. So we enabled that
the enterprises to reuse that with a platform. And like that, we
started and we rolled that out. We added remote support and things
like that. And over the years we realized, okay, now that that’s
working even if there are a lot of challenges and you know that,
we’re not there to say it like that. In a way they say, “well AR
is kicking off in a way that everybody’s using it with the CRM.”
But the next step for us, and now coming back to the initial point
about “what does it mean to build the operating system for
enterprise AR?” is, to have a platform is one thing. But there
are many vertical specific requirements today. Whether you go into
medical or pharma, where you have regulation, or you going to
automotive or aerospace. The scenarios are completely different, and
the requirements are different. And we as an AR startup — and that’s
the same for other AR startups — we cannot solve all of these
problems. We are technology companies. So this is why we said we take
our platform, we make it modular, we add SDK and API to make it open,
and building that operating system so that partners, system
integrators, solution providers, but also enterprises on their own
can build solutions they need. And this is exactly what companies ask
us for: to have a solution where they can customize their stuff,
where they can integrate into their solutions. The funny thing is we
always think as AR company and you mentioned it, it’s a community and
it’s cool. We are all so cool in AR, but in enterprise business,
they’re not going to change their infrastructure and their
architecture and the systems because of AR. It has to happen the
other way around. And that’s why we’re doing what we’re doing.
Alan: What do you mean by that?
“They’re not going to change their infrastructures is off
because of AR. It’s the other way around?”
Dirk: What we tried is to offer
our solutions and then go into the enterprise and say, “Well,
look, here it is. You have an AR solution and now you need to
integrate into that. You need to provide your data and all of that.”
But enterprises have existing infrastructures and architectures. They
have a CRM, they have an authoring system, they have a PLM system,
they have a CAD tool. They have all of that. So AR is just one little
thing, and AR has to integrate into those kind of systems, not the
other way around. No company is going to change a lot of their
infrastructure or their CRM or their CAD tools because of AR, that’s
not going to happen. And that’s exactly why we deliver it as an
operating system.
And now we allow and say, okay, for
instance, we connect to Siemens PLM Team Center. So that means a user
who already works in something like Team Center does not need to
leave that environment. Our platform kind of runs in the background.
It is used for publishing, for helping with the technical stuff,
making that easier, like building a tracking configuration. All of
those things are integrated into our platform, but the author can be
at the used environment which he or she already knows. So it’s not
about learning all of that new. It’s just to have kind of an AR layer
on top of it. And that makes it much easier. And that’s — for me and
that’s for us — the key to make AR scalable and also to have finally
acceptance on a higher level. Because the technology is available,
it’s not a problem at all. We can realize all of those things the
enterprises need. It’s much more about the acceptance among IT people
and also among the frontline workers.
Alan: Interesting. Yeah,
somebody mentioned on one of the other podcasts that we don’t have a
technology problem anymore. We have an adoption problem.
Dirk: Right. That’s how it is. I
fully agree on that. Well, look, as I said, we can realize it. I
mean, give you an example. We had one little gap in the publication
process. You could do everything. You could upload your data, you
could add your AR content. And then at the end you could publish. The
only thing you could not do was to build your own tracking
configurations. So it means you need to tell your AR system and your
camera and your algorithm when it sees a machine. What does the
system have to place on top of the machine? What is said to be
overlaid here? And that’s called a tracking configuration. That had
to be done by an engineer or by a software developer. We closed that
gap; we created a tool. And with that tool, everyone can do that,
without any kind of coding skills. You just use your existing data.
You compare it where now expressing a very simple way and that’s it.
So it closes the gap.
That supports what you just said. We
don’t have a technology problem. We can fix all of that. Now also
with machine learning on top of it, which improves the tracking and
the usability of the whole system. But don’t forget that, especially
when we talk about the frontline workers, the actual users, then we
have to talk about change. And you can just go and say, “Well,
look, you don’t use your hammer anymore, now we give you an iPad and
that’s it.” That does not happen from today to tomorrow. We have
a new report we worked on, why so many POCs or trials get stuck in
that early process and do not proceed to a deployment, to a rollout,
what we all want to have in the AR community. And it’s about change,
it has nothing to do with the technology, maybe with the usability.
But at the end it’s change, and you need to tell the people why they
should use it, what is the advantage, and how can they use it?
Alan: All right. So let’s unpack
that. Why should customers use this? How do they use it? And then
what are they going to get out of it?
Dirk: Talking about the why,
they’re the obvious things, which we repeat over and over. Talking
about enterprise use cases, you can reduce your downtime, you can
improve your first time fix rate, reduce your error rates. There’s so
much more than only that. I’m thinking about that you can really
improve your customer experience. You add a lot of value across the
product lifecycle, things you could not do before. And I said when I
explained that example, when we did it for Range Rover, that that
showed it clearly. Take it from the consumer side: you buy a new
product — let’s take a coffee machine. So you unpack that. What do
you want to do? You just want to use it. You don’t want to flip
through pages and read how that works. And it’s an awkward
experience. It takes you forever until you get that coffee, finally.
Alan: Don’t mess with people’s
coffee.
Dirk: Yeah. [laughs] My coffee
machine did not work in the morning. I’m lost without my coffee! But
nobody wants to read the printed manuals anymore. If you find them,
then they’re outdated. It takes a lot of time to understand what’s
written there. You don’t want to have that anymore. And let me add
that to the why. I think it has a lot to do also with what we know
from consumer side. Even that many people say, “Oh, well, you
know, but it’s enterprise, and enterprise is okay to be boring.”
No, it’s not. We are used to Netflix. We’re used to iPad. That kind
of experience. And then we go to work, and then we have to use the
old tools which do not provide anymore what we need.
Alan: It’s crazy, right? We have
these beautiful UX/UI for everything in the consumer world. And then
you go into these enterprises, and it looks like Windows 91.
Dirk: [laughs] Exactly! And
that’s the problem. People don’t want to have that anymore. You come
from home and you go to work and you have to go into another world.
And that’s not acceptable anymore. And that’s what we see, that
transformation happens. And also we think about the next generations.
If I see my son — and he’s eight years old — and I showed him the
computer and I gave him a computer mouse and he looked at me and
said, “Dad, what’s that?” And I said, “that’s a
computer mouse, for your cursor here.” And he just left. And I
said, “What?” He said, “No, I don’t want to have that.
I will use my iPad.” That’s a change in behavior. And that’s why
I say we need to bring the consumer grade experience to the
enterprise sector. And AR helps us a lot to do that. Now coming to
the how, well, that’s a bigger part. I mean, we could do several
episodes here about how. But let me quickly given all of you about
that. I think it’s a lot about how you can get started. In the past–
the last years, many enterprises kind of wanted to get started with a
huge thing. Everybody was thinking about that rollout. How do you
decide that if you were a manager in a large enterprise, are you
going to decide and say, “I will invest 500k into a new
solution, where the ROI is not proven or I don’t know what’s going to
happen exactly. I see there is a potential, but I don’t know exactly
how it’s going to work. And I have to report that.” Nobody’s
going to make that decision. And that’s, I think, one of the problems
we had in the past.
Alan: I always tell the story
how when we started meeting with customers, when we started, they’d
go, “Great! This is amazing technology, I love it, it’s really
mind-blowing! Who else is doing it? What are the ROIs and how much do
they cost?” And we would say “Nobody else; we have no idea;
and a lot. Still want to buy it?”
Dirk: [laughs] Yes, that’s
reality. Not augmented; only reality. That’s exactly how it is,
right? And those are the questions that we have every day. It’s not
enough anymore to have some interesting logos on your slides and say,
I work with ABC. No, you have to prove it. How many users do you
have? How did they do the rollout, as you said it? What is important
is you need to do it in steps. You need to start with that with a
trial. Go do a POC–
Alan: Ok, so what are the steps?
I’m a new customer. I own a factory. You guys come in. What’s step
number one?
Dirk: So step number one is, for
us, we start with a little package and do a usually a scoping
workshop, where really to check what are the requirements? Is there a
company who just wants to test something? Is there a company who’s
already thinking about a bigger solution? And then specifically, what
is the use case? I really want to see a use case when we get started.
It does not need to be any complex thing, but just one thing. And
then in the workshop, we define “Okay, that’s the use case.
That’s the way we do it today. And then we have KPIs. And that’s the
way we’re going to do it with AR.” And then we can measure.
That’s the key. So and that gives the enterprise a small package
where they can define what they do together and with the front line
workers that’s important. It needs to go bottom up, not only defined
by the management, because that will be a failure. And then you have
that scoping and then you have the measurement. You can say, “Okay,
this is how we did it before, and this is how you do it with AR. And
now we can see that’s successful. Does it make sense in that use case
[garbled]
?” So once you have that, you go to the next step and
you go into a POC where you can really test that internally, within
several machines or several products and really start to scale it.
Not only in a test environment, but in a real world scenario, which
makes it different or even involve your customers and get the
feedback there. And then that’s the second step. And then to come to
the third step where you finally roll it out. Take all the experience
you have from step one and step two, take that into consideration.
And then finally, you’re able to roll it out. Because between 1 and 3
dictate a lot of different steps.
Step 1, it’s more about figuring out
how it works, maybe already involving the IT a little bit and say,
“Okay, so what kind of connections do we have here? How do you
install all of that?” And step 2 is then completely focused on
the business case itself. “Who’s going to use that? How is that
compared to the current solution? What is the customer or the user
feedback?” And step 3. It’s a tough part to roll it out. It’s
about security. It’s about integration. It’s about all of those
things. And that’s a completely different story. I mean, it takes you
a month to really get the security done and get approved as a vendor
with a large enterprise to fulfill all the requirements.
Alan: So this scoping workshop,
POC rollout, is this a 12-month rolling timeline? Or longer, or
shorter?
Dirk: It depends a little bit on
the enterprise side. There are some there, they’re faster. But there
is some that need more time. So bringing that into a timeline that
the scoping workshop and the results out of that. That’s something we
do. Let me say maximum 60 days. That’s the timeframe. What we also do
is, we do the scoping but we also give them the product to work on
that, to get a feeling. What can they do? How can they create content
on their own or with our support? So we really train them, we have
our own education program where we train them. So our promise is up
to 60 days. A company knows how to create an application on their own
and they create the first application on their own. And then, for the
POC, you can add another 90 days, because that has to be prepared. So
60 or 90 days for the first two parts. And then rollout, I would
definitely plan a six month time period. That’s what you definitely
need to prepare all of that. It’s a lot of communication and a lot of
things to check internally. Month in, you can finally roll it out.
And even then, you probably do it in a few countries or with a few
users before you do that globally. It’s a big thing, to do that you
have a lot of different requirements. Might think about things like
GDPR we have in Europe, which you do not have. And the same in the
States or in Asia. Think about Asia and China. You cannot just go
there and host a server. It’s just not possible. We tried that and
luckily we have a partner like Bosch, our investor and partner,
Bosch, and they helped us with lots of stuff there, because you
cannot just go there and say, “OK, here we are. No, let’s get
started.”.
Alan: I guess not.
Dirk: No, it’s not. It’s not
that easy. It’s possible. And the market is growing fast.
Alan: What are the KPIs? What
are the measurements that we’re going to measure? What are some of
the metrics that that you would test against them? And is there a
cost to phase 1? What are the costs associated with phase 1, phase 2,
phase 3?
Dirk: Yeah. With phase 1, it’s
usually a wrench between, let me say, $10k and $20k. POC and rollout
is a little bit harder to say. POC depends on what exactly is the
scope of it. Let’s say there is between 20 and 50. But in rollout,
it’s hard to put in numbers, because that’s very specific and depends
on what kind of solution you want to have. But of course, then it
goes into licensing and subscription. That’s tough, but it’s also a
50-50 plus then, to have that in numbers. So KPIs you asked for. So
that’s something that– the interesting thing is the enterprises that
they know their KPIs, they know exactly what they have and what they
need. The typical things, what you have is error rates, first time
fix rates, down times, when you think about the manufacturing sector
especially. That’s a very clear KPI. But of course, there are many
more if you look into the training sector. How do you measure
knowledge obtainment? How can you measure knowledge transfer? When we
think about the skill gap and the workers gap in the manufacturing
sector. And they have the big issue that they need to transfer the
knowledge from the people getting retired to the young guys.
Sometimes you have really hard numbers. Like what I said before, they
reduce time to downtime or the error rate. And sometimes it’s a
little bit more difficult to say, “OK, how can we measure things
like knowledge transfer?” and we have to define that together
with the customer. It’s very important to find a base, to compare
that. We had a customer and they did it a completely different way.
We asked, “So how did you measure the success?” and the guy
who led the project, he said, “Well, it was simple. I asked my
team what they think, and the team said we’re super happy, and it’s
better to work than before.” And I said, “Okay, I
understand. But what are the metrics behind that?” And he said,
“No, there are no other metrics. It’s just the team is happy and
works better than before.” I said, “OK, that’s an
interesting way to define a KPI.”
Alan: [laughs]
Dirk: Well, this shows you we’re
still early stages, right? We’re building KPI frameworks for our
customers. But it’s not that you have one framework which you can use
for all kind of customers. It’s just [garbled] word right now. I had
a meeting last week with a large enterprise in Florida and for them,
other things were super important. For them, it was important how can
we automate the content creation? How can we customize the UI, that
it fits to what we have already? We want to integrate that. We don’t
want to have that our service technicians have to get used to a new
UI or something new. So that was for them much more important than
other matters. They said, “Well look, that’s what’s for us now
super important, because that helps us to integrate the system and to
show the employees and the workforce, well look, that’s a good tool,
use it. And then we will look for the other KPIs once they’re using
that.” And I like that approach because as I said, it’s about
change. So the first step, you have to make sure that people want to
use it. What do we have right now in 70 percent of the cases, we
create a tool which is very helpful, but maybe workers or technicians
use it once a week, but they don’t have that frequency that they use
it every day. So it does not have the value it’s supposed to have. It
really depends on what is your goal. And then you can define your
KPIs.
Alan: Now, you’ve run a number
of these in this protocol. So this 1-, 2-, 3-step process — scoping
workshop, POC, and then roll out. How many of your customers are in
the rollout phase now? Or maybe already rolled out?
Dirk: Several are in the rollout
phase. The interesting thing is — and that’s also the sad thing for
us enterprise AR guys — is that were not allowed to talk about all
that stuff.
Alan: [laughs] I figured that.
Dirk: That that’s one of the
most difficult things. But we’ve already rolled out. So like the
biggest rollout we had was an application, which is called Daimler
Rescue Assist. And I think it’s probably one of the biggest rollouts
in AR history, because it is available in 27 languages and for all
Mercedes cars since 1990.
Just to give the background quickly,
why it is so big: it’s an application for first responders. And when
they come to an accident, they have to figure out where are the
critical parts like battery wires, fuel tank, things like that. And
they have to make a decision, in seconds, where they cut into the car
if there’s someone inside. And before, they had to do that with PDF
instructions, and you can imagine that’s quite tough if you have to
compare that. You have deformation in the car. Then you have to cut
into that car. That’s not really a good decision-making basis. That’s
the reason why we have that application. And Mercedes and Daimler,
they wrote that out globally, together with our partner Bosch. So
that’s one of the big examples. And that shows also it’s possible to
do that. And the others, we have currently between five to ten
rollouts in preparation, communicate more about that soon. There are
different companies from the chemical sector, but also from the
energy sector. And what you can see is, it’s completely different.
There is not one rollout. I mean, you have some things that always
happen, it’s like security as the major topic–
Alan: What are some of the
things that are always consistent? From what I’m hearing, you’ve got
one major rollout, a couple more coming. So, we really still are in
the earlier stages of this technology. And again, it’s an adoption
issue, not really a technology issue. But that is only recent that
the technology has caught up to our visions.
Dirk: Yeah, absolutely. I mean,
I think it’s a lot about once we have seen the first five, 10, 15 —
that will be the trigger. I think many enterprises are waiting,
they’re ready to go, but they’re waiting to see others. And then they
will quickly turn into the–
Alan: Fast follow.
Dirk: Yeah, exactly.
Alan: I live in Canada; we’re a
country of fast follow. We watch what America does, and we wait for
them to make their mistakes, and then once it’s proven, we jump on.
Dirk: That’s like the Apple
strategy, right? [laughs]
Alan: Exactly. Never be the
first; always be the best.
Dirk: Exactly. But that’s what
we see. They want to seed, they invest, but they don’t want to make
the decision about the really big deal. And then on the other hand,
also, it is a big deal to do a rollout like that. And if you have
like several thousand of users involved there, a lot of questions you
have to answer. And that starts. You have to involve like branding,
who is going to host your app? Who’s going to do that? What our
security requirements and all of those things. Where are we going to
start? Which country is the first one? What kind of devices do we
have? And don’t forget about by large enterprises.
We often have old operating systems,
whether it is about Windows or even tablets and stuff like that. So
you cannot really compare it. There are some basic patterns. Security
is one. I would say less like 70 percent of the securities that
that’s basically the same large enterprises. Then there are some
specific rules. But basically that’s the same. So what you have done
that. And you know how to handle that and you have the right tools
for it, then that becomes easier. But the other things that’s really
different because organizations are different.
Alan: So what are some of the
considerations? Let’s just run through a list and we’ll just compile
a big list and we’ll put it in the show notes for people so they can
at least have a list of considerations. Right now I’ve got security,
branding, hosting location devices, device management, integration.
Dirk: Right. Countries. That’s
something we have to check because there are different devices we
already mentioned. Those are the well precise and the league terms in
terms and conditions. So that part is very, very important. How do
you do that? And then it depends also on is that an internal pronoun
or John Roland where he goes to the customers? That’s also completely
different. That’s what we what we have right now, where the customers
are invoked. Then you need to think about, okay, how do we provide
access to customers? What kind of systems do they use? So it becomes
kind of a narrow level of complexity on top of that. But you learn a
lot in different units, have different requirements. And then I think
one of the biggest parts when you when you get started with that is
to find the right people for it. You really need the right people in
these large enterprises who know how it works. That’s for me, always
a major point. We come from our AR world and we think we know
everything and then we come to an enterprise, and the first thing we
have to do is we have to listen. Because we have to learn about, how
does that enterprise work? What are the structures? How is the
organization organized, and what were the specific things they can
tell you? There are so many different things that they have to look
for. And then it’s about customization. If you can, define steps for
the rollout. But that does not mean that it really fits for all
enterprises. You need to have an individual plan for an enterprise to
do a rollout, that it fits. Otherwise, it’s not going to work.
Alan: All right. So what are
some of the challenges that you guys have seen that you’ve had to
overcome, over the time that now, is just part of your process? What
are some of those challenges that people getting into this [can
expect], that seem to be general challenges?
Dirk: So I think one of the
biggest challenges is what I mentioned earlier: the way we get
started. Instead of saying, “well, look, here is a license,
here’s the product. You get a training: now get started. Create
applications and roll them out.” That’s what we thought we can
do. Just more or less, you sell a CRM and enterprises install it, and
they get started. But we’re not selling a CRM. We’re far away from
selling CRM. When you buy a CRM, it’s more about, “which one do
I need,” but not whether you need one. With AR, it’s about, “doI
need AR: yes or no?” It’s not about which kind of solution, the
first step. And most of the enterprise, they even don’t think about
it. And most of the enterprises I see, they even don’t know about the
different solutions or vendors. That was one of the biggest
challenges for us to learn and understand. What is the right model to
get an enterprise started? And large enterprises, mid-market,
industrial market — it’s just the same. There’s no difference from
that point of view. So that’s why we came up with the different
stages, right? Scoping workshop, POC, and then rollout. That helped
us a lot to learn together with the enterprises, and take the
learnings, improve it, and then do the next step. Enterprises are not
ready to buy a license right away and get started with it. There are
[some]
, of course, but that’s not the typical way to do that.
Alan: So, you’re a platform.
Dirk: Right.
Alan: You’re also a content
provider, and you’re also a service provider. You’re kind of having
to do everything, is that what I’m hearing?
Dirk: Yes.
Alan: That will change as this
technology becomes more adopted, and people can start to realize the
benefits of it. But I think where we are right now in this
technology, from what I’m hearing from you, is that we’re still so
early that we still need to handhold every customer through the
process of deciding that, “this is something we want to do,”
and then trying it. “Yep, it does do what we say it’s going to
do,” and then, “okay, it’s time to roll it out.” But
the companies are still ill-equipped to do this themselves. Selling
just the CRM system is not possible just yet, but it will be in the
future, as more and more case studies come out and people are able to
say, “okay, I want that,” right?
Dirk: Exactly. That’s what you
describe is reality today. The interesting thing is that applies for
all our startups. If you have investors, and investors — of course
— look for recurring revenue. So, you shouldn’t say, “okay, we
will sell licenses and nothing else.” But in these — and that’s
not only for AR; I see exactly the same in IT, in block chain, in AI
— we are not at a point I can just sell the licenses and the
customer takes it and does whatever. We need to have the education
and the service part. There are many, many enterprises. They ask me,
“what kind of person is that, who would create the content? What
kind of skill set do they have need to have? Whom do I need to hire?”
That was, for us, the trigger, and we started now with our own
education program, where you can get your own modules, and companies
can book that, and they can train and educate their own people to do
that, [and ensure] that they have also the same level of education
and same level of knowledge. We need the services until, as you said,
we reach a point that everyone knows. But even look at the big
systems. If you look at SAP, or even a CRM — take salesforce,
something like that — you need a lot of customization. You need a
lot of service to implement that in the right way. And AR is no
different.
Alan: It’s interesting you
mentioned that, because this is one of the reasons we started XR
Ignite, of which you’re a mentor, as well. XR Ignite was started as a
community hub to foster the relationship between startup studios and
developers, and corporate clients. Because what we realized — same
thing you just mentioned — is that the content creation is a
separate thing from the platforms, and venture capital invests
typically only in products and platforms. But what they’re missing is
this sweet spot of content creation. And one of the things that we
realize is that independent developers and small studios, they’re
going to start to get acquired, because companies — exactly what you
said earlier — they’re going to say, “who do I need on my team
to build this content? Oh, I need 10 people with Unity skills. Well,
I don’t have time to hire 10 people. Why don’t I just acquire this
little studio here that can do what we need, or work with them for
six months? Test them and then acquire them.” I feel that
there’s going to be a lot of these small micro-acquisitions of $5-,
$10-, $20-million acquisitions, where teams are just basically
acquiring talent.
Dirk: Absolutely. And you can
see that already, that companies are thinking about that. How can
they solve that? Do they have own teams mostly? No, they don’t have.
They have like one or two–
Alan: No, and it is going to be
hard, because right now, there is a shortage of this talent. And if
you project out two years, when all of the companies all of a sudden
wake up, the talent pool is not going to be much bigger, because
schools and universities and colleges are not teaching it yet. And I
think we’re going to find this issue right across the board with
exponential technologies, that our current school systems are
ill-equipped to retrain people. How do we train people fast enough to
become very, very proficient in creating this technology, so that
companies can start to either acquire studios or startups and really
build them in, or hire them?
Dirk: Absolutely. And that’s
what we need. And it’s different often, right? I mean, content
creation is something… especially when it comes, now, to the two
big things we do: repair and maintenance in areas, in training
environments, and in the manufacturing or medical sector. You take
your CAD — people know how to do that part — and then you add your
AR layer on top of that. But then it already starts, and is okay, how
can I do that? But I think is there is a shift from 2D people to 3D
people today. Content creators, developers. I used to build 2D
interfaces. What we have in our mobile phones; tablets. But now,
handling that in the 3D space and make that a great user experience
and guide the people and help them to use it? That’s a completely
different story. And we don’t have that. There are almost no people
— think about UX designers — being able to think in a 3D space in
that way, especially in the enterprise sector. There may be in the
consumer area, but in the enterprise sector, there is a lack of
people.
Alan: I agree, and I would
punctuate that with the fact that ARKit — which is Apple’s framework
for developing AR — and ARCore — which is Google’s — majority of
the really interesting, robust content that’s coming out of ARKit and
ARCore is consumer-driven. It’s games, it’s experiences. Even if it’s
a business application, it’s usually some sort of marketing. These
applications — while also very, very valid — designing enterprise
AR is not as sexy as designing a video game. It’s more lucrative, but
less sexy. So you’ve kind of got these game developers over here that
are making amazing games in three dimensions. But there’s a skills
gap in between the enterprise and the gaming world.
Dirk: Yeah, you need to know how
enterprises work. You need to see how one of these frontline workers
is using the tools. And that brings me back to how we started with
AR. I remember some of the apps, where we showed a car mechanic how
to change a tire. I mean, that’s not going to work. We need to see,
where does AR make sense? Where does it provide value? And that’s
something that you have to combine that with a good user experience,
because it has to be interesting. It has to be also fun, even though
it sounds weird for an enterprise sector. But it is like that, coming
back to what we what we discussed earlier; you need to have that
Netflix and Apple experience to have a good work result. And that’s
completely different to what you do for games. You can learn from
games, right? Not to underestimate that. You can learn a lot from
games: how to keep people interested, how to keep people at a level,
how to provide information and transfer information. There are so
many different things. See a firefighter: how they hold the tablet
and use it with their thumbs. Maybe they have gloves even. Then go to
a car mechanic. This is a completely different story, and that’s what
you have to understand, too.
Alan: Oh, man, there’s some so
much here. So let me ask you, on a personal note: what is one problem
in the world that you want to see solved using XR technologies?
Dirk: Well, that’s a great
question. There are a lot of things I see, but I think for me, even
that it sounds a little bit global now. What I want to have is I want
to enhance my work environment. I don’t want to be tied to little 2D
screen. That’s where I see the biggest impact of VR. Give me my 3D
work space, where I can define my environment and I can take it with
me. And I always have my office with me wherever I am without having
a little 2D screen.
Alan: And then your office does
not need to be limited to a desk with four walls anymore. You could
create an an an environment… like, I like working on the beach in
Bali. So, every time I put my headset on, I’m in Bali on a beach,
working on a beautiful desk with 100-inch screens around me. The one
thing that we don’t have yet is the ability to have a keyboard and
type in the area, which is crazy. One of those startups that applied
for XR Ignite, actually, is working on that. They’re working on a
keyboard that allows you to see your fing– well, not your fingers,
but see your keys in VR. And it allows you to type in a more
comfortable way in VR.
Dirk: That’s definitely
something we need. And the things I’ve tested so far were not really
usable in the way I want to have it. So right now, I think we only
can enhance our current desktop office, but we really cannot replace
it. But what you mentioned is exactly right. I mean, you see that
often: the keyboard is really a key element in the way we work.
Absolutely. Yes.
Alan: Yeah. And actually,
Logitech introduced their their three dimensional pen recently. I got
a chance to try it at Augmented World Expo. And then there’s another
company called Massless. And basically what they’ve created is an
input device, looks like a pen, but it’s tracked in 3D space. So now
you can draw in three dimensions; something like Z space, which is a
computer that allows you put on really inexpensive glasses and see
three-dimensional things come out of the screen. They’ve also got
this kind of stylus effect, where you can use a pen to grab things
and bring them into 3D space. So I think the user input devices — be
it a keyboard, or voice, or gaze even (looking at something), or a
pen — we’re only just starting to figure these things out.
Dirk: Yeah, we’re early stages.
On one side, we want to reinvent the whole thing, which makes sense
because we need it. It’s a different environment. On the other side,
we also need some elements which we’re used to. Right? I mean, that’s
always funny. Many people also said, “so you tested a lot of
variables; a lot of AR glasses. So, why do we always have the typical
rectangular screens in that 3D space?” I said, “well, look
it’s just the beginning. But also keep in mind, if you learn
something new — if you have to get used to something — then maybe
so it’s good to have some things you are familiar with, right? And a
rectangle screen, with an X at the top right corner that lets us
know, okay. I know how to deal with that. And that’s also a good
feeling.” So it’s a transition phase somehow, where we have to
learn. And it’s interesting that you mentioned the stylus. Yes — I
played a lot around with the Apple pencil. Everything is digital
today. And I came back to that and I didn’t want to use paper,
because I don’t want to have another notebook in my backpack. So I
use that. And I have to say, it’s really a good way to work. I can
imagine that it helps us to work in space.