Reimagining the factory floor as a metaverse workplace

Plex CTO Jerry Foster explains how the metaverse could enable factory line workers to perform their duties remotely through a combination of virtual reality technology and high-speed connectivity.

Episode Transcript

Dave Essex (00:03):
It's the ERP con fab. I'm David Essex industry
editor as search ERP ismanufacturing the next industry
that will be transformed byremote work. It's not as
unrealistic as it sounds,advances in virtual reality,
factory automation and remotemonitoring could soon make it
possible for line workers to dotheir jobs remotely, thereby

(00:26):
expanding the talent pool andhelping to alleviate the
shortage of skilled labor. Italked to Jerry Foster, CTO and
Founder of Plex, one of thefirst multi tenant SAS ERP
vendors. But why he thinksremote factory work is a good
idea. And the steps needed tomake it a reality. Plexus PR

(00:47):
firm pitched a conversationabout an idea I really hadn't
heard before, which is howvirtual reality can be paired
with automation and remotemonitoring not only to create a
factory floor of the future, butto allow factory workers to work
from home and to have trueremote work. Can you describe
what you're envisioning and kindof fill in the details,

Unknown (01:09):
maybe I'd set up just a little bit of foundation to that
there's I think there's a fewcomponents in play two of them,
that's already in play. Thefirst thing is having, you know
cloud based software, like PlexMBs, which runs in any browser,
with an internet connection,that gives you the mechanism to
not just control your operationsremotely, but also monitor your
entire enterprise from anywherein the world. So it gives you

(01:32):
total visibility, regardless ofthe geography, the name of the
plant, so workers can do theirjob remotely, as if nothing had
changed from when they were inthe office. And decision makers
can continue to make criticaldecisions because they had real
time, full visibility into everysingle one of those plant
operations. So as an example, wehad one of our customers during
the pandemic, they had to sendeveryone home as most people

(01:54):
did. And half of their plantswere using cloud based software,
like Plex, and half were usingon premise software, the plants
that were using cloud basedsoftware continued running as if
nothing had happened. But theplants that were using on
premise software, it reallystruggled, they couldn't
communicate, no one knewcorporate didn't know what was
going on in those facilities.And so that's the power of a

(02:15):
cloud based remote situation andscenario. So you think about,
okay, in that scenario, they'vesent all the workers home,
except for the workers on theline, right? Because they still
needed someone there, pressingthe buttons on the factory
floor. So envision thisscenario, envision being able to
send those workers home also.And they would go home, and they
would put on their virtualreality goggles. And they would

(02:37):
interact with a digitalrepresentation of their machine
or their work center, whichwould send a signal real time to
its physical twin, locateddozens or hundreds of miles
away. And it would actuallybuild the part or do whatever
thing that they were trying todo as if they were standing
there in that actual physicallocation. So using this concept
of digital twins and, andwhatnot, we have the scenario

(03:00):
where everyone could go home andwork from home. But for the line
workers, we'd actually have toadd that additional capability
of using those VR goggles andthat digitalization of the of
the assets, so they caninteract, interact real time,

Dave Essex (03:12):
how close are we to having this be real? And how
many pieces are in place? Orwhen do you think there'll be in
place?

Unknown (03:20):
I'd like to say, okay, September of 2028. Right? When
you talk about a situation, likeI just mentioned, there's a
number of technologies in play.And so each one of those are
going to happen in a step bystep basis, right. Some of those
things will require implementinglike 5g at your facility, some
will require you to upgradeyour, your network or your

(03:42):
machines. And some of that willrequire implementing a cloud
based mes. So there'stechnologies along the way that
have to happen and are happeningright now then being able to
pull all those things together,is what's going to actually
enable the kind of the Nirvana'sscenario that I just envisioned
a few minutes ago, I think justbased This is purely finger on
the wind kind of guesswork, Iwould expect to see functioning

(04:03):
prototypes, probably in the nextfive years.

Dave Essex (04:06):
I can sort of imagine the benefits, but I
guess I need to ask what theywould be like, for example, you
know, we know there's a shortageof skilled factory workers in
that manufacturing work isgetting very technical. Do you
think this will help with maybethe upskilling issue or just the
issue of getting the skills intothe company because those people
can be remote?

Unknown (04:25):
So there's a lot of aspects to that question. So let
me let me hit a few of them asfar as the benefits and biggest
benefits, just the ability toweather unexpected disruptions,
right? Can you imagine someonein a boardroom, some
manufacturing executive in 2019,that says, Okay, our strategy
for 2020 is to figure out how tosend everyone home and not go
out of business. Who would havethought that we would have

(04:47):
actually had to come up with astrategy like that for Cloud
enables that very thing. Sothat's a huge benefit of this
remote type workers to weatherthose unexpected disruptions.
Then on the employee side, ofcourse, they can continue to do
their job. There's a Work lifebalance component not having to
commute and, and not having topay for gas, environmental
benefits, reduction in fuelconsumption, wear and tear on

(05:08):
our cars and roads, mentalhealth benefits. So there's all
these benefits surrounding theworker. And then you touched on
the shortage in the workforce.And I think there's some real
advantages there as well, forthe for instance, one advantage
is you no longer have to hirefrom a local talent pool, you're
not restricted to just those whocan commute. And then to take
that a step further, it breaksdown a lot of barriers, you

(05:30):
might have potential workers whoare restricted because they
might have physicaldisabilities, or they can't
drive for one reason or another.Well, they can use a computer,
they're very adept withtechnology, suddenly, they can
be part of that workforce andhelp you fill that gap. So
there's definitely somecapabilities there. I would also
say, another component of this,which is fascinating to me, is I

(05:50):
mentioned a second ago, thiswork life balance components.
We've had workers here at Plexwho during the pandemic moved
home to care for ailing parents,we had one worker who sold their
home and bought an RV andtraveled the country for a year
and a half with their family andlittle kids and didn't miss a
day at work. So you got thisability to I think, to give
employees a bounce that makesthem fulfilled, and happy

(06:13):
employees are more productiveemployees, which is part of the
solution to this workforceshortage. And then the last
thing you touched on is theskills gap. I think there's some
interesting things there. Whenyou talk about the skills gap,
I'm not sure that this remotework, addresses it and might
actually exacerbate the problema little bit. Imagine the case
of like a manufacturing purchaseagent. If we send that person

(06:33):
home and they're going to workremotely, there's really not a
whole lot that they need to betrained on to do that their
scenario at home is very similarto their scenario at work, they
have a browser and theirinternet based software, and
they're doing the same thing.But now imagine that Nirvana
situation with the line workergoing home? Well, it's not just
them in a physical machine, it'svirtual reality goggles, and a

(06:56):
digital representation and awhole technical component over
and above their normal job, it'sdefinitely some training that
has to happen there. My guessis, it's going to take time to
kind of bridge that gap. And Ithink actually, we're gonna see
a lot of new workers beingattracted to this, you know, the
the kids who are growing up witha pacifier in one hand and a
tablet in the other, I thinkthey're going to gravitate to

(07:19):
that sort of skill set and thatsort of capability.

Dave Essex (07:21):
I know that virtual reality and augmented reality
are also increasingly used andof great interest for corporate
training, how will this help arethe training getting people up
to speed, the upskilling, andall that, there's one

Unknown (07:35):
thing to tell someone how to operate a 2000 ton press
that come, you know, comeslamming down on a piece of hot
molten steel and shifted into apart, just telling them how that
works, and actually putting themin front of that machine and
kind of the safety issues thatare involved with a newbie on
that kind of scenario. And maybeif you're trying to practice on
actual material and the wasteinvolved in that. So if we could

(07:56):
actually set up an augmentedreality scenario where they have
a press in front of them,they're actually working on
that. And we have an actualrepresentation of what happens
in real life in front of them,they can become very comfortable
with how the you know, the lookand feel and the actions of
moving your hands and how itworks is so much different than
just reading it on a piece ofpaper or having someone tell it

(08:18):
tell you so the step from paperto doing it in real life versus
in augmented reality scenario,to real life is much smaller,
and that ladder situation. So Ithink there's a huge possibility
there are potential I shouldsay, and that sort of situation
for sure.

Dave Essex (08:32):
Now, what are the risks? I do certain operations
in a manufacturing facility thatare harder to control remotely?

Unknown (08:40):
Yeah, for sure. I think there's definitely some risks.
How do you address unforeseenissues in the plant? Anyone
who's worked in themanufacturing facility for any
length of time knows there'salways those Oh, crap situation
of what do we do about this? Andyou just, you can't even plan
for them? Because you didn'tknow they were gonna come? You
know, what happens when awindstorm knocks over a stack of
pallets? all over the floor?You're AGVs can't get it, you

(09:02):
know, get around it. Who's thereto clean that up? What about,
you know, electrical issues inthe plant? How do you solve for
those? You know, sometimesthere's those complex situations
where you just need a human tostep back and say, Oh, this is
what we need to do. So I thinkthat's a risk. I think we've
already touched on the skillsgap is a bit of a risk. I also

(09:23):
think there's a bit of a risk.And this might be off the beaten
path a little bit, but we gofull remote, and I've even
noticed this at our customersand plus even a little bit you
risk a lack of Team cohesivenessof camaraderie. And I think
there's an intangible tangiblebenefit there. Even on the
factory floor, I work shoulderto shoulder with these people. I
get to know my co workers. Ilearned about their kids during

(09:45):
lunch break and I have theirback. There's a sense of
protection there with yourcoworker that you get when
you're working shoulder toshoulder so I think there's some
some risks in losing that. Thereare definitely operations that
are harder to control remotely.And how do you it's how harder
to automate, you know, a multistep process or complex
production scenarios on changingour tooling from one job to

(10:08):
another can be difficult. So Ithink there's definitely going
to be a long tail to some ofthose some of those use cases,
getting those ironed out. Ithink there's some basic use
cases that we'll be able totackle. But some of those
complex ones are just going tohave to be solved over over
time.

Dave Essex (10:23):
Do you have any advice about how manufacturers
can try to get their mindsaround deciding which functions
can be done remotely? Maybe it'stoo early to say, but are there
certain characteristics ofthings that can be done remotely
versus ones that can't? Andthat's how they can sort of
analyze the situation as theydecide whether they even want to

(10:43):
try to do this remote work,

Unknown (10:44):
right? It's a journey, I think it is kind of hard to
say there's lots of learningsalong the way, you just don't
jump into the deep end, right?You like, what's the first step
to this? For a lot of companies,the first step is just
installing or implementing onselling but implementing some
some cloud based software justto get their users used to the
fact that, hey, I can do at homewhat I do at work. So maybe

(11:05):
having accounting work from homefor a couple days a week,
there's very little risk there.And it's a step towards, hey,
this is a remote scenario that Ican play with. And then I think
the next thing to think of iswhat are the economies of scale?
What are the most basic,repetitive tasks that we can
automate? And then then weautomate it, we put robots and
Cobots in place, then we trainour workers to manage those

(11:26):
robots remotely. And so that'sthe next step. I think there's
just a step by step processthere. And you start with the
easiest, what can you have haveimmediate success with and go
from there?

Dave Essex (11:36):
Would you call this a Metaverse application?

Unknown (11:39):
I would, but I hesitate doing so because the term
Metaverse right now is so muchbaggage. And it's getting beat
up, you know, in the news quitea bit, and rightly so. And
there's so many obnoxiouscommercials and promises. But
that's because that's where weare in the innovation cycle.
We're in the hype section. Andthat's what you expect in the
hype section. I've seen it overand over and over with all the
technologies that we've beenthrough. And that's where we're

(12:01):
at right now. So I expect atsome point as we enter into this
trough of disillusionment, partof the hype cycle, eventually
we're going to kind of love alot a little bit and we're going
to actually see some real usecases take hold that are
actually productive in theirimplementation.

Dave Essex (12:16):
What about the technical hurdles? We've talked
a bit about some of the peopleissues, and I think you
mentioned 5g earlier and sometechnologies. Can we kind of
revisit that? Sure. What aresome of the technical things
that people need might need toput in place to really make this
feasible?

Unknown (12:30):
So I think the thing to remember here is when you talk
about Metaverse, the metaverseitself is not a technology. So
you don't say I'm gonna goinstall the metaverse. It's a
collection of technologies,right. And those technologies
have to work together. And weknow that goes sometimes. But we
have to get these complextechnologies like digital twin
and 5g, virtual reality, spatialtechnology, even artificial

(12:52):
intelligence a bit. Those thingsall have to work together. And
those are the things that makeup the metaverse. So there are
definitely some technicalhurdles there. And getting those
things in place are going to bekey to helping make this happen.
So you know, setting up adigital twin or mate or
implementing 5g in yourfacility. Each of those is a
step towards making this more ofa reality. I think there's also

(13:13):
a couple other infrastructureissues, high speed connectivity
everywhere. And that's becomingless and less than issue. But I
live out here in the boonies.And we just got high speed
internet last year. I mean,yeah, last year, last summer. So
there's some places that youwant to send your workers home,
but when they get home, they'relike, I can't, I can't do what
you need me to do. So that's athat's technical hurdle. The
power grid that bulletproof,that's a hurdle. So some of

(13:36):
those infrastructure and utilityissues have to be resolved as
well,

Dave Essex (13:39):
I see. So I guess the typical factory, and I
honestly, I haven't been in onein a little while, um, they tend
to have their own backup power.So they're not really at risk
within the factory of losingpower for any length of time.
Right.

Unknown (13:52):
But the home workers might be, you know, how many
help workers have have, youknow, generators or backup or
that, you know, we just had awindstorm last week in this
area, and not many people outhere for days. And if they're a
critical worker in thatinfrastructure, then I need to,
you know, what am I going to doabout that? Am I going to
subsidize generators for all myworkers? How am I gonna think

(14:12):
about that?

Dave Essex (14:13):
I'm wondering about the smaller steps that maybe
companies can take along the waythat would provide some
immediate value, maybe beforethey put together the whole
vision, I would say

Unknown (14:23):
it's very similar analogous to the entire digital
transformation process. When youlook at industry 4.0. There's
what a dozen technologies atleast, and variations of those
that you could bring in house orinto your factory to help
improve your processes. Youdon't just drop all those
technologies on the floor andsay, All right, or industry 4.0
transfer transformation iscomplete. You know, everybody's

(14:45):
awesome. And it doesn't workthat way. It's okay, I need to I
have a scrap problem here. SoI'm going to install a quality
management system. And I havesome machine downtime issues. So
I think an IoT system would helpthere so your path to industry
four I know it's one step at atime. And it's the same way
that's going to happen with thisremote worker Virtual Reality

(15:06):
Metaverse sort of thing, right.And we kind of talked about this
a little bit already. If Iimplement 5g at my facility,
that's gonna go a long waytowards helping connect sensors
and the physical devices in adigital twin environment, every
time I upgrade a piece ofequipment, so that it has
network connectivity, as opposedto some of the World War Two era
show machines, I still see itfactories, every time I upgrade

(15:28):
a new machine, that gives meanother step towards that
process. If I implement a cloudbased mes, that's a huge step,
because I've suddenly given allmy work is the ability to work
from home. Um, so there aresmall steps. And each one of
those is just a piece of thepuzzle towards the bigger
picture. What is

Dave Essex (15:45):
Plex specifically doing to make this possible? Do
you have this enabling of remotefactory work? Is it is that part
of your software development oreven your roadmap for your
products,

Unknown (15:55):
three questions that we'll look at in just a little
bit differently. I would say, asa cloud based software, enabling
remote factory work isinherently part of what we
already offer. So the ability todo work remotely, is an
advantage that Plex customershave had for almost 20 year over
20 years, actually. So actually,our roadmap focuses more on
improving not only improvingfactory operations, but enabling

(16:17):
those automation capabilities,the things that will support,
remote factory engagement andeven higher levels. One example
would be our demand casterproduct that helps forecast
inventory levels with accuracy.So in this case, demand caster
will actually point out when amanufacturer might see an
increase or decrease in demand.And that allows them to staff

(16:38):
appropriately and respond tothose spikes or those decreases
and that automation. And that'sall automated. So that's the
decision makers can then makethose decisions on where they
need those workers on call ornot. So we actually build
capabilities into the softwarethat enable that automation,
which in turn provides aplatform for for remote work. I
would also add though, and thisis kind of intriguing, I think

(17:01):
some of that last mile typestuff required to build out a
true Metaverse is much morehardware and controller and
network centric. And that wouldfall into the wheelhouse of our
parent company, RockwellAutomation. They just acquired
us. And they are already layingthe groundwork for this. They
have I should say we now becausewe're part of them software,
like emulate 3d in the FactoryTalk twin studios products.

(17:24):
Those are right in the in thewheelhouse of virtual reality.
And the metaverse and they'realready laying the foundation
for that. So we think that thecombination of Rockwell and
their hardware centriccapabilities and Plex coming in
and filling a niche there withthe software side, you basically
have a kind of a wall to wallsolution there that is that is
moving very quickly towards avirtual lifestyle factory.

(17:48):
Metaverse, whatever you want tocall it,

Dave Essex (17:49):
you know, just said you've been in the cloud for
more than 20 years now thecompany has been around for 30
years. I'm wondering, have youalways had the multi tenant SAS
Type of cloud? Or did it evolveat any point from let's say,
single tenant SAS or hostedcloud some of the other kinds of
cloud that are still aroundtoday that, in my view aren't
really the full purest form ofcloud, which is multi tenant

(18:12):
SAS.

Unknown (18:13):
Right, exactly. So there's a bit of an interesting
story there. And the shortanswer the question, the short
answer to your question is, itwas shared multi tenant from the
very get go. Because back then,there was no cloud, private,
hosted or otherwise. So thingslike single versus shared tenant
really didn't even exist, webuilt our own private cloud, one
of the first enterprise systemsin the world to do so about

(18:36):
2000 1001. Right there, alongwith Salesforce, I think they
did at the same time, although Ialways say we probably did it a
couple of months earlier thanthey did. So we were very much
on the on the leading edgethere. And the interesting thing
there is obviously, if thecompany has been around 30
years, we didn't start in thecloud, even before we built our
own. Our first system was a Unixsystem with dumb terminals,

(18:58):
where the dumb green screenterminals. And so we were used
to this notion of, we would makea change on that central Unix
server, and all those terminalsthroughout the plant would
automatically reflect thatchange. And then in 95, right
when the company started,officially, we broke off from a
14 company started our owncompany, Windows 95, was their
client server was all the rage.So we built a client server

(19:21):
system. And it was awful.Because this concept of making a
change of a single location wasgone. We had to implement
whenever we made a change, ithad to be installed in every
single PC at every singlelocation in the entire plant.
And we were like, This is crazy.This is really hard. So in 2000,
when the internet came along, orat least the commercial, you

(19:43):
know, view that we were able toget, and we looked at a browser,
we realized, oh my gosh, abrowser is basically a dumb
terminal. All it is doing is isrendering what you push to it
from a central web server. Andwe're like, oh my gosh, we have
our central deployment. Writtenmechanism back. And not only is
it plant wide, it's worldwide.So even though it didn't spawned

(20:06):
from a single tenant, it spawnedfrom this notion of the central
piece of control that allowed usto make change at all of the
entry points around the aroundthe plant. And so we sat down in
99 2000, to build this. And we,we not knowing how forward
thinking we were at the time,because we didn't know we didn't
know, we built a shared multitenant system and the core that

(20:26):
architecture is still in usetoday, because it's so powerful.
And so advanced or fourth time,

Dave Essex (20:31):
I think an analyst or two that I've read Recent
years have had noted, the ironyof this kind of cloud is like a
throwback.

Unknown (20:38):
Yeah, in fact, I've got a slide when I give
presentations, if you thinkabout every 20 years, 20 or 30
years, back in the was the 60s,you had these timeshare
computers, right, you had thesecentralized servers, and the
VAX, and all that stuff, andthen we moved to client server,
and then 2030 years later, we'vemoved to cloud. And now where
are we moving kind of back tothe edge, right, we're talking

(20:59):
about edge computing and edge AIand, and so there's always this
there's this 20 to 30 year cycleof centralized at the edge
centralized at the edges is kindof interesting.

Dave Essex (21:09):
Regarding multi tenant SAS, especially for
manufacturers, I've been reallygiven the impression by some of
your competitors, who I, it'sfair to say have struggled to
get their manufacturingcustomers to move to multi
tenant SAS. So I'm thinking ofSAP primarily. And there's a few
others, they held back by theircustomers need to continue

(21:29):
running the factory systems,including the MES as the
manufacturing execution systems,and probably the ERP itself on
premises because those systemsneed to be a highly customized
and that maybe it's easier tointegrate them with the factory
automation systems and the othermachines. How does Plex view
that issue since your productsdo run on multi tenant SAS? For

(21:53):
years?

Unknown (21:54):
It's a great question. So I'm going to split that into
into two parts of that. Okay,the first part is this needs to
be highly customized, right. Soremember, you've got decades of
ISVs, making a living withcustomizations, and then make a
killing, helping you through thepatches and upgrading. So of
course, there's going to be verystrong pushback in the SAP
ecosystem against the platformthat sort of cuts out that

(22:15):
business model, right. Plexdoesn't have that baggage. Being
cloud based, we're versionlists, every single customer we
have is on the very latestversion of the software at all
times. So they're always up todate don't have to worry about
patching and upgrading. But thetruth is, there are sometimes
customizations needed. And wehandle that three ways. We have
a standardized REST based APIthat our customers can access,

(22:38):
we have our own internalextensibility option where
customers can build their ownapplications and write their own
sequel to support it. And thenthe customer can petition us to
add those customizations intothe product, which we at times
do judiciously, right. Beinghighly customized, I think, is a
throwback to different models.And we've addressed that through
our extensibility because it isrequired. I just wanted to touch
briefly on the integration witholder systems, it is a valid

(23:01):
point. But it would say youknow, we have a full set of
API's and integrating withfactory automation systems like
the older systems, and like Isaid, we have a full set of
API's for those situations,whether you want to go in the
direction of say an oldercorporate ERP, or maybe down in
the direction of the shop floorwith older systems. But what our
customers do and is really nodifferent than the SAP world is

(23:24):
if if that older system can'tcommunicate with our API's, they
would simply write a piece ofmiddleware and orchestration
solution there, where thatsolution actually integrates,
however, that older equipmentneeds to be on one side and
connects to our standard APAPI's on the other. So we can
communicate that way with theolder equipment on the shop
floor.

Dave Essex (23:43):
The other day, SAP at Tech Ed announced a low code,
no code tool, which is somethingthat a number of ERP vendors
have introduced. Do you have oneare you thinking of introducing
one,

Unknown (23:54):
we've had one since about 2007. And it was actually
the second solution that Imentioned. And the previous
question about how we addressthose customizations, it's
called Vision flex, allows ourcustomers to build their own
screens, completely native anapplication without having any
coding capability requirementsto that. And they can use
standardized data sources thatwe built to populate and run

(24:17):
those screens. Or if they want,they can go a little bit step
further. And they can write somebasic sequel to access the data
and the way that they want needto support that screen as well.
So so we do have thatcapability. Well,

Dave Essex (24:29):
I guess it's great that you've had that experience,
because I have to say, when Iwhen I hear these announcements
from the other ERP vendors, myjournalistic skepticism hat goes
on. And I think really nontechnical users are going to do
that. Really, this is going tomake them do that. And it's
going to open up the developmentworld. It sounds too good to be
true. And I'll admit, I haven'ttried to use any of these tools

(24:51):
yet, but it sounds like avariation on things that I've
heard off and on over the yearslike, this is process management
workflow tool. holes,diagramming and all that.

Unknown (25:03):
Yeah. So there's there's always a gap between the
promise and the realization, orthe claims and the realization.
So I don't know what they'reoffering. I know our local tool
is pretty powerful, becausewe've been using it doing it for
a long time. But there's alwaysa little bit of technical, you
know, skill required andunderstanding how to build the
screen and how to move thewidgets around and how to
connect them to a data source.So there's some, there's some

(25:24):
capabilities there that youneed. So but you know, like I
said, we've been doing it for awhile, and hopefully have ironed
out a lot of those bugs. So who

Dave Essex (25:30):
would those types of users that are doing so they're
not the business only personwho, which is what it sounds
like for some of these tools?Are they like, not a programmer,
but they have some pretty goodcomputer chops? And I'm afraid
of computers? Is that the kindof person who does it?

Unknown (25:47):
Yeah, if it's business analysts, it's people in the
accounting department who want atweak to an existing report and
add a couple of fields that theydon't have. Sometimes some of
the some customers do have an ITgroup and they want to see
reports or screens in their ownway in their own format. And so
it's across the board. Yeah.

Dave Essex (26:06):
Well, Jerry, this has been an interesting
conversation. And I thank youfor your time for joining us
today. Every day, thank you.