Digital Twin Your M...

In this webinar on Digital Twins, we'll cover:

  • Why dashboards won’t solve your remote monitoring challenges

  • Applications of digital twins for remote working

  • How to get started with...

Transcript

In this webinar on Digital Twins, we'll cover: - Why dashboards won’t solve your remote monitoring challenges - Applications of digital twins for remote working - How to get started with... so welcome to this webinar on digital

twin your most productive remote worker

and we'll look at how to detect and

respond real-time events when your your

team can't be on-site or at the physical

assets the focus all today is wide

dashboards won't solve your remote

monitoring challenges the application of

digital twins for a mud monitoring and

also how to get started with your first

remote monitoring digital twin the goal

of my presentation is to help three

types of people and organizations the

first is those that have already started

on remote monitoring and the digital

twin journey actually as a result of

having to move their control room to the

bedroom as what has happened over the

last couple of weeks with with with many

organizations the second audience is

really those who have already started

implementing some remote monitoring but

it's not really helping to tell the

organization's win key business events

are happening so they've started doing

it but not really that effective and

then lastly the the goal of the

presentation is to help people

organisations who are not going to go

back to the old way of doing things and

that have recognized that the next

normal or the new normal require more

responsive and remote digital

capabilities going forward

so it's not for the laggards I'm Peter

Funt scopic I'm the the chief digital -

an officer at XM Pro and I've worked in

both engineering and IT and I'm probably

version 1.0 of what is called a digital

engineer as we starting to see

traditional engineering skills are now

converging with with digital or IT

skills I have a passion for solving

complex problems and being actively

involved in the industry internet

searching with a number of initiatives

including digital twins and I've

contributed articles to Forbes through

the folks Technology Council but I've

also had the opportunity to kind of work

firsthand with digital twins and look at

how they impact operations and

maintenance in a number of the

industries that we work in

and I take some prior trying to a

few licks or of those examples with you

today so one of the things that we've

seen with the covert 19 pandemic is that

it's impacted different industries in

different ways and but for most of us it

has actually changed the way that we

work day to day

Jarrod's Patara was on the interview

with such an Adele a couple of weeks ago

when he said we will never go back to

the way that we that we worked in the

past and in a different thought leaders

and analysts agree that you know the way

that we thought we were going to work in

five to seven years time from now is

actually how we started working two or

three months ago we all expected to work

from anyway that capability but we never

thought it would be forcibly

fast-tracked to us and this combined

with the maturity of emerging

technologies like IOT cloud age AI

machine learning all of that is creating

a perfect storm for event based remote

monitoring and digital twins so before

yeah so what we covered today is really

why dashboards won't solve your remote

monitoring challenges we see a number of

customers trying to start there and with

that I might like to share with you how

we stumbled on this event based remote

monitoring drones while working at a

fortune thin super major oil company in

in it was actually in 2014 so we've been

around this space for a while and one of

the number of challenges that they had

was trying to address how to schedule

their truck rolls for cruise around

compliances inspections and that that

work impacted the production yield on

their on the oil field so they try to

manage this with spreadsheets and bi

dashboards and and the challenge is that

the decisions that they made were on

were done on style data that was

actually sent to a title like with batch

by CT else I think a lot of you are

familiar with that kind of thing

and you know extra fact some of the data

lag by as much as 24 hours

the dashboards required a skilled

engineer to interpret and look at it and

decide on what actions need to be done

and and when that engineer one person

suddenly had to take extended paternity

leave no one knew how to interpret the

the dashboards itself so decisions were

typically made on on information that

was that was out date and didn't really

reflect what was happening around

operations and maintenance events and

there was also no way to link those

actions and those decisions with his

work actually being done because there

was no way to track the dashboard versus

the actual work they needed to do so we

need to try to build that into our BPM

or workflow engine but we realized that

the sheer volume of streaming data

needed to do this in real time kind of

needed a whole new approach so we

created a remote monitoring digital twin

built on event-driven architecture and

I'll spend a bit more time on that and

it's really for real-time operational

intelligence and we call it event

intelligence so it's about that

real-time what are the key events that

are happening so you know when we got

this right we saved them eight million

dollars in six months and we also

reduced the truck rolls for those

services by eighteen percent so again

looking at some of the challenges that

we see with bi styled dashboards kind of

my two main ones is the one is access to

real time data and the other one is

closing the loop so we briefly touch on

those and remember kind of the data like

way the oil and gas customer trying to

put all the information and then with

the reporting on the challenge where

that is that you are using bi style2

tools to do that way we first stole and

then we analyzed and what real-time

remote monitoring requires is actually

access the real-time streaming data

where you analyze and if you you may or

may not store the information so as it

happens you need to analyze it very

quickly gotten actually graded or really

and just by the way um what what we've

seen

is the tip of the iceberg or the easy

part is what we store the harder part

and what you really need and the thing

that that that that the sink many IOT

projects or remote monitoring projects

is the fact that there's expectation

that those dashboards can do that the

the top part what what we forget is that

there's a the the mass of the body

actually resides under this gardener as

a nice way of explaining this and this

is from a gardener slide but really if

you look at bi and all of those we store

and then you analyze versus what we see

from continuous operations intelligence

or event intelligence where it's kind of

embedded in the business process and as

you can see we analyze as it happens

which is reflected by the process it

gives us situational awareness it's

always on pushes the information to us

but it also orchestrates why do we need

to do when that happens prescriptive

action and the whole idea is that we

close the loop so it is a process that

you can build around this and by the way

you can store data if you want to then

which is reflected there a better way of

looking at this as well is also from a

remote monitoring perspective time

relevance of data is really really

important so when a key business event

happens and that might be that we can

predict something's going to file or

that are that are operational that you

need to make operational changes or

something like that when that event

happens the information that we have

then start appreciating very quickly

once we know something is likely to

happen or happening in some instance we

have you know the information starts

depreciating within a second a minute an

hour or a day which you see at the top

and when it's completely there's nothing

I can do about it anymore

a slightly different time cycle and we

have this window of opportunity that sit

between the two and that's the time that

we have to respond so we typically want

to know when these key events are

happening and just remember these

revenue yellow dots I'll get back to

them a little bit lighter but those are

key things that we want to know and what

we've seen is business intelligence is

when the sand was settled I can go back

and analyze you know below the line when

it's had

and and I can understand how many file

is that I have why did I have it from a

feint intelligence that's the sign

that's still in flight it's something

that I could potentially still impact

and that is what the benefit that we see

with remote monitoring so if you get

this right and this is an example of a

mining customer that started out with

Excel and B I'm not going to spend too

much time on this but every single red

cross that you see is when equipment

filed that cost the impact of this was

literally millions of dollars and 41

hours of downtime on equipment and I'll

show you a little bit more of that

environment in in a minute but these are

the conveyor belts underground 1.2

kilometers underground 80 kilometers of

conveyor and these are failures on these

conveyors trying to manage it with Excel

and bi

only getting to it you know once a week

once they've implemented the XM Pro

monitoring capability from a digital

twin for the for the for those

underground conveyors you can see the

impact that it's had because we can do

that now at scale the monitoring twin

can do this every two seconds we can

give them advance notice and we can tie

it back into you know are we actually

going when we tell you something's

happening are you going to do it and

this is actually the the environment so

this is a under this is a borer

underground 1.2 kilometers underground

and the stuff that or a camp that comes

out of this goes onto a conveyor belt

which when this project initiated and

you know we took this initial snapshot

of in the in the early version of of

wind and this is looking at conveyors

you can just see the red and yellow dots

you can see a lot of problems and some

of this resulted in in in high downtime

the the second part is what we took a

coupla days ago so you can see how it is

actually changed in terms of being able

to monitor this better the interesting

part of this is the person who knows how

to do this their reliability engineer

that can look at the recommend that can

look at the event provide a

recommendation if someone has been there

45 years lots of experience and it's

actually someone who is at risk at the

moment in terms of

the curve at nineteen virus so that

person is working from home he's been

able to do all of this from home for the

last couple of weeks so you can see the

benefit of having moving the control

room to the bedroom but bringing that

remote monitoring digital to

incapability into it and the from our

perspective there's a big difference

between what you do on a dashboard this

is what we refer to as an event board

and which is this digital twin that

represents where are the key events that

are happening and I'm not interested in

the speeds and feeds and I don't need to

know what the tank level is I just need

to know if there's a problem with that

tank and what do I need to do when that

happens so that's kind of the first

issue that we have with bi style

dashboards and and things like that

where the second part is really how do

we close the loop and I'm gonna talk a

little bit around that by showing you

some applications and by what I mean by

closing the loop and with looking at

some applications of digital twins for a

my cracking I wrote a article in Forbes

not too long ago around how to hire the

right digital twin for your company and

will provide this link to you in the in

with the recording that we send out

later but when you look at a digital

twin it's very similar to hiring a

person in your organization and there

are key things that you want is to this

this thing to do so kind of the the

identified three major jobs that we hire

digital twins to do the first one is

just tell me what's going on that's the

status twin the next one is you know

tell me what I should be doing digital

twin which is our operational

interacting it's kind of giving me

advice and an interactive and then the

last one it's kind of the Oracle in the

organisation it it's the tell me what if

scenario simulation predictions

capability that we see with digital to

ontology to go and read that that

article explains it in a lot more detail

but I thought I'd show you some examples

now ignore what you see on the bottom

right which is the recommendations

because we just can't help ourselves we

have to kind of shove it but for the

rest what you see is a typical status

TWEN what is happening right now - and I

can get a view on on different aspects

and this is built in XM Pro this is a

slightly different so this is an example

of an operational twin so again this

tells me where certain things are

happening so if you look at the previous

one I don't see any see any real red

dots or events or anything like that I

just see I know what's the status now I

see there are problems in certain areas

like at the pumps and and things like

that and this is also recommendations

associated with each one of those so if

you look at the bump the the dot in them

over here you'll see there's a

recommended recommendation associated

with that specific pump and also there's

some information around how good are we

at resolving this so kind of connecting

back to our weave are we fixing this and

how good are we and you'll see this is a

constant pattern that we repeat in these

remote monitoring digital twins this is

that same one with a little bit more

simulation capability so this runs a uni

a Unity gaming engine at the back and

you can drill down and fly in and in the

prison in the demonstration a little bit

later I might get to that and so the the

but this is the real area that you know

we when you look at a simulation twin or

something like that and this is what's

actually done by one of our partners

they are on the call so this is a

simulation when that does casting

guidance for blast furnaces is their IP

there's some really smart models and

things that sit at the back what you see

it over here this is actually a unity

based model that can move up and down

I'm not going to get into the specifics

of how this works but with this I can

run scenarios I can do all sorts of

things again taking real-time monitoring

data and providing costing guidance -

and operate

who needs to make decisions the the

benefit of this again doing model tuning

and and and setting up new scenarios can

be done remotely by people who are not

on site who might actually be sitting in

their bedroom again again this is done

by one of our partners if you're

interested in in in this application and

they've specialized the engineering firm

who does this happy to if you contact me

happy to share their contact details

with you and you can explore how this

could apply to your business so in terms

of closing the loop which is the last

part so what we often see is there there

are multiple systems that provide and

that's one of the challenges with

dashboards is kind of how do we close

the loop I can see information um I have

all these systems that give me alerts

but what I'm really trying to do is is

kind of connect that up to an action and

with those events that you solve

triggered earlier those red dots and

yellow dots this is a recommendation for

each of those that you can set up and it

shows the real-time information that

that triggered that event a

recommendation from that engineer who's

been there 45 years and it's now working

from his bedroom and some interaction

with and putting out work instructions

and you can even bring in work order

information and some of the other

components the way that we set that up

is we wire that up and this is again

something that you can you can give the

digital twins some intelligence some

rules that it needs to follow to figure

out how to trigger those key events by

by setting up rules and this is fed from

an XM probe data stream if you look at

the XMP idea stream it is that thing

that sits under the water when you look

at the iceberg the big mass or body

master that then that was sitting there

so again looking at that mining

organization closing the loop for them

in that they also not only understood

what the key events or but how often

those we triggered how people

responded to them how long it took to

actually resolve some of them and again

this changed significantly from when it

started and it moved from a dashboard

Excel environment into a monitoring twin

and that has got feedback loop so I just

lastly just want to touch on how to get

into your first digital twin and I'll do

that by explaining a little bit in terms

of our methodology but then take you

into showing you how you would actually

do that using our application now we've

created this lean digital twin canvas

it's bold it's it's basically a business

model canvas that we've adapted for a

lean digital twin it's kind of the one

page view if you had to add one slide

that you can present to your executive

to get funding for this digital - and

what would you put on that and the

numbering that you see on here it's kind

of the sequence that you that you put

this together so if you started on a

digital twin is and that's what's what's

the problem that I'm trying to solve we

often see people start with I have this

data what can I do is it it's kind of

the wrong end what you know you need to

focus on what is the problem that I'm

trying to solve who will be my customer

will be the user of this remote

monitoring for example what is the value

prop that for this and and what does the

solution compile consists of so what are

we going to include in that what are

some of the challenges that we might

have and and as you can see the business

guys the key matrix that we are going to

measure to make sure that this is a

success how can we establish a baseline

and then what will this cost and what

integration challenges maybe have in

what way do we need to get the data from

this is really the job description for

that digital twin that you want to hire

once you've got that it makes it a lot

easier to go into how our standard

approach of developing digital twins it

always thoughts with what is the problem

that this from work monitoring digital

twin will be solving for you and kind of

we've divided it into three main

components again we need to look at what

are the

we need to get the data stream so once

we know what the problem is we've gone

through you know what are the key

metrics and what are the leading

indicators that we're trying to get to

we they need to see so how do we get

that data streaming into our rules and

you can see again this we've broken it

down into easy chunks of work to do how

can we then define what those and those

rules and recommendations how would we

actually give that to someone in the

event board with notifications and and

also integrated into existing systems

like EAM and ERP and some of the other

things and this is kind of the this

standard way that we look at doing

things so kind of the three steps that

we follow in doing this in the first

step is we will create the data streams

and I'll show you that in a minute

we then set up recommendation rules and

then we create the visualizations and

the front-end for for this so that's

kind of the three simple steps in terms

of how to get started so with that maxsa

going to show you a live example and

just want to get that across a layer and

I'm just going to open up X and press

suite

this is the standard Exim Pro module so

we'll start off as a as I mentioned with

the data streams so this is where we get

the data into our digital twin now that

we know what the problem is and again

always thought with what is the problem

that you're trying to solve so you'll

see we refer to these actually as use

cases what are the use cases that you

want to do and it might be I just want

to confirm that everyone can see my

screen so oh cool got confirmation yet

you can see my screen so this is where

we bring data in and these are built

around specific again use cases or

applications and if I look at something

like maybe a center that's in traffic or

pump that we saw the red dot that we saw

on there I'm not going to run through

the whole thing here but the whole idea

is how can I set up a data workflow for

my hundred pumps or 200 pumps using one

data workflow I can I bump the data in

from all of them I'm getting it from

time series data or maybe some protocol

and I'll do that by listeners drag and

drop and again a whole library of this

there are separate videos that cover

this on our Learning Channel

and so urge you to go and have a look at

those if you're more interested in this

context where do we give the mic and

model from a side B and other

applications like that and then how do I

set up some analytics for this and

wrangle the data how do I bring in and

then go and call a predictive model for

example in this instance and then in

what's what's nice with this is you can

also change predictive models and this

one will predict that it's likely to

file an enema go and run a completely

different predictive model and again all

of these X and Pro comes boxed was a

bunch of AI and machine learning

capability already the drag-and-drop

anja so you can buy kit into into your

data flow that's really where you want

to use this transformations is just how

you wrangle the data and and do

calculations and things like that

functions is where

Bly and maybe first careers or something

if I want to look at bearing vibration

and the last thing is I want to create

some actions I want to see nowadays and

missus I want to send it into the XM Pro

managing work orders and then or into

ASAP or Maxima or in for that doesn't

matter what the system is that you're

sending into but I also want to run

these recommendation rules so again a

bunch of capability but what I wanted to

get across here is and you know all of

these are pretty easy to set up the

whole idea is you can construct the the

veins that flow or the blood that

through flow through the veins of your

digital twin through this this is kind

of like the blood that that that flows

through that and again pretty easy to

set up the different systems to be able

to talk to and create the logic that I

want to feed the information for my

real-time monitoring to and so this now

tells me from a monitoring perspective

what's happening to this bump so that's

the first part the second now that we

know that and we've sent it on to our

recommendation rules so there's some

data coming out of this pipe okay

actually go and have a look at that

quickly before we look at the actual

applications and screens itself so if I

look at the recommendations and actually

just going so this is all the

recommendations and there are different

ways that you can manage and I'm

actually going to resolve that one

because I want to show that these fire

in terms of real time but before but so

I couldn't manage them at a high level

but what I'm what I'm what I want to do

is actually show you how we set them up

so if I look at that discharge pressure

on the pump the one that created that

red dot was actually this one but it's

only got one rule in so I will actually

have a look at the second one which has

got an additional rule so you can see I

can bold rules on this incoming data

remember this

so my digital twin

blood Turing the flowing through the

veins is now sending me in some

information that it's collected along

the way that information I can now use

to set up a recommendation to say well

when this condition occurred make it a

red condition or a yellow condition or

whatever it is or medium or high or

whatever the case might might be angry

eyes some response that I need I want to

know what with work instructions and and

integrating it into our existing

workflow cyst or ein system and adding

resources like video clips on how to fix

this bump but in order to do that I need

to set up a rule and this rule the data

that you see coming out of this is

actually the data coming out of this

pipe over here so that is how we connect

and that's the part that's often missing

in dashboards you can't do that you can

I'll show you a dashboard in a minute or

if input but without this capability of

bringing data at scale being an analyzer

bring it down to what you need and this

board is is the part that you need to

set up these rules and recommendations

to make your event you know this will do

in any event when that will tell you and

by the way you know this could also be

that I want to say that the flow rate is

less than the feed right so you can

build some really interesting rules and

this is actually maintained by that

engineer working from home who's been

there 45 years he knows what the rules

are and also and we have also invested

in building this rule the rule discovery

through AI topic for a separate webinar

but in this instance it's actually set

up by and the recommended actions by

that person has been there 45 years a

great way to capture and make a digital

to an intelligent using the knowledge of

people who's been there a long time and

understand what they're looking at and

translate that into and codify that into

into a real tongue digital twin so this

is the the recommendation now that we

have that let's see what it looks like

if I go

look at what that recommendation

actually does and I'll quickly go in and

look at so if we look at a at a at a

mining plant for example and a

processing plant on a mine what you'll

see conveyors there's there's a pump and

there's the equipment I can see there is

an issue with that I can see how long

we're currently taking to do some of the

things but they will see that there's

actually discharge exception on this

pump but I'd actually like to understand

more around the pump pump so this is

capability that I now have from a remote

monitoring digital to an I'm sitting in

my bedroom and I can see the information

if I'd want to drill down into some of

the other organizational data sources

like in this instance the historian

that's got the history of the last 20

years of this pump collected I don't

need that right now if I have lost seven

days I've got enough and and I could

actually even have the work order

information from the I said from the

same system whatever display on on this

as well and yeah I can see some of the

key metrics but I can also look at what

the recommendation is around that pump

and what you'll see in this instance

there is the data that triggered so the

data that's coming through from my data

stream is not has now triggered that

condition that we set and that condition

these are the actual values that created

it

I can then create and a comment or

instruction I can then assign it to the

work order management system either

automatically or depending on how the

organization want to use these

recommendations so that's one way of

kind of closing the loop and why a

dashboard doesn't give you that

capability doesn't give you the ability

to see to back to command and control

and give the feedback and if we look at

one or two other examples

on this it could be as simple you know

it doesn't always have to be fancy

pictures like the previous one that I

showed you it could be some something as

simple as now these are all my conveyors

and I can see the health score again I

would get alerted when some of these

with the red condition and I get SMS

with the yellow condition and I get an

email and I would know what what the

recommended action is and what the data

was that would triggered it I could

drill down in a similar way a different

view of what a digital twinness it's not

always a fancy 3d CAD model or something

like that but if we if we were to do

that in extra fact let me show you an

example of just a close of an example

that is more sophisticated so this is a

water treatment plant and this digital

twin actually runs a a unity gaming 3d

rendering of the real-time data so it

will see this is my real-time plant I've

got some bumps and things here and I can

navigate and go into and actually and if

I just like different views then I can

actually draw in and you know what

you'll see what it does it actually

shows me the condition and there you can

see the real-time data stream so again

there is action agent in here that will

anything from smart contracts on through

to things like unity models and we can

we can make the digital the remote

monitoring digital to an really really

intelligent with different capabilities

but the real but the real time data and

get through the recommendations in

everything that you see is still exactly

the same and serious got a question for

me which is probably a leading to into

the next section on our webinar so that

is the the way that we put this

twins together we start with the data

stream within create recommendations

that we want and conditions and then we

create the actual application and the

way that we put these applications

together if I just quickly go back to an

example like this conveyor acid one it's

pretty easy to to create these

applications inside example I have

rights so I can edit this one I

shouldn't because our precise theme will

be will will have to go and and roll it

back right if I update the version but

this is how I build and design so I can

layout I can drag things on yeah I can

make areas like this a widget and they

reuse that widget so you'll see we've

got widgets and then you can very easy

to actually build it out but you don't

start here you don't start with the

dashboard you start with what is the

data stream what are the recommendations

that sit behind it and then how do I

want to visualize it and show it to

people

so with that time for any questions that

you may have on on this so leave a few

minutes for for questions and answers

and if you're interested well you know

we can help you

using our three-step process we can help

you to build a real-time and digital

twin monitoring twin that can connect

the data and kind of give you that

unbiased right timing information so

that you know what you're looking at and

you can track the leading indicators for

key events with that so that any

questions

there is one I'm just going to open up

the question box the world virtual

coffee be provided so I think one of the

interesting things with these digital

twins and thank you for that question so

I think one of the interesting things

with these these digital twins is you're

only limited by your imagination in

terms of what you can put together and I

always use the example of traditional

status twin is something if you think

about your car it is like your fuel

gauge you can see what the level of the

fuel is and you need to be an

experienced operator if you've been

driving for a number of years you'll

know exactly what what that means when

it's high or low and whether you can get

to your destination that's kind of

typically what what we see with bi star

which is it's it's that when we look at

some smarter solutions

you know the lightest versions of the

Tesla's if I put my GPS if I put my GPS

coordinates in I can actually the the

call will will look at the charge that

is got it will determine whether it can

actually get to that destination if it

can't it will flag that it's an event it

will then look at so this is

intelligence that you can build into

twin like this it can then look at what

are the charging stations on the way

where is an open slot in one of those

charging stations and do you want me to

book it for you obviously that will

optimize everything and once you once

you there and answer your question

around the virtual coffee well what you

can actually do is if there's a

Starbucks nearby you can actually at the

digital to an order your stop your

Starbucks coffee and and and donat while

your car is being charged I know it's a

little bit of a different kind of take

on that but yeah so I think the other

some of the other

around years how long does how long does

it take to create a digital twin like

this of typically do we have a six week

program from and you can find

information on our website as well we

have a six week kickstart program where

right from the start we you have a

digital twin delivering value is is

typically around a six week window

depending on on access to data and how

much you want to do in-house so one of

our objectives is to try and your people

to be able to do this yourself

the there's a question around issue of a

three step approach applicable for

building a supply chain digital to an

absolutely we've built a digital to an

around supply chain solution for a

really large multinational organization

and it follows exactly the same thing we

need to understand kind of what the key

objectives were and the instance it was

sort of inventory planning then you know

what are that the the data real-time

data that they need and how do we get

access to that and how do we get it into

a shape by using a in the into a shape

that I could use it to create

recommendations and those

recommendations around short-term

inventory planning and then lastly how

do we now show that is someone and how

do we tell them to do something and how

do we integrate and I actually use ASAP

as the backend and the question so I'm

just running through them quickly have

you have you integrate your platform

with clamp point models or models yes so

the a number of of different ones and

one of the interesting digital twin

knots if you call it technology or

platforms that are emerging or for

example the Microsoft digital twin which

is really a BIM model

and integrating the information out of

that yes so managing the data flow is

the key yes absolutely so it's not about

the visualization it's and thank you for

that question

it is about how do I get the data from

the real time sources how do I reduce

the lag how do I make sure that I don't

necessarily store but I just I analyze

and then decide if I want to store but

that is kind of the biggest challenge

that we see with us with this question

is how do we handle potential false

alarms great question one of the things

that we've done initially and again

we've been around the spice for a while

we and we made a number of mistakes as

we as we went along this one of the

first things that we try to do is take

that data stream and push it and have a

predictive model and then push it

directly into a seppie or Maxima

whatever the case might be ended up with

false positives and every single time

you send out someone to go and have a

look at a machine and they get there and

there's nothing wrong with it you lose

credibility so that's one of the reasons

why the recommendations are and one of

the things that you'll see on the

recommendation is was this

recommendation success did it resolve it

yes or no and the other option is you

can actually say was this a false

positive so that you can go and tune

your model and and there's always at the

moment there's it's not decision

automation it's decision support so

we're not one of the downfalls as we

started off trying to do one of the

challenges we had was trying to

automatically grade those work orders

yes we can is it a good thing typically

not until you really trust information

you fine-tune your models and you know

that they're working well so that's how

we how we handle false positive alarms

and there they are are different ways of

also doing it or discovering that

through some IR technology that we're

using it or digital twins for machine

manufactory manufacturers possible

absolutely so if you and we are actually

working with machine manufacturers that

make really interesting components in

that

actually have IOT capability now built

into the component I deployed tens of

thousands of this and that information

gets fed and they are selling the hard

way with the software that the remote

monitoring event one for their software

so yeah

in terms of of doing this for actual

machine manufacturers absolutely and

there are different OEMs that use our

software to create solutions for their

hard way which they ship with the

software great question on how do you

how do you work with real time data and

this probably the last question that we

have time for if there are any more just

please send them on and we'll get back

to you but you mentioned that your

platform works on real time data how do

you capture these data points in real

time like in the mining example they

might be network issues or data points

what I didn't explain and apologies for

that again there's a separate video on

how there's in more detail and around

the data streams is that that data

stream could actually be deployed oh

sorry distributed across multiple

environments it could be a hybrid that

you might do some of the initial data

acquisition on-premise or at the edge

depending on you know what what term you

use for that and that you might do some

of the pre-processing there it may white

if there's network of connectivity

issues and send some of the data which

you then analyze for that this could

also be deployed on-premise so it

doesn't it's not necessarily cloud by

solution it is it could be deployed

on-premise it could be deployed in the

cloud or it could be deployed as a

hybrid to deal with network connectivity

and that example that I showed where the

partner created the solution around the

blast furnace they actually have

connectivity issues so we had to in the

header history Mike provision for how do

we get data in when we when we actually

don't have network capability and that's

through edge device and doing that part

of it so great question thank you for

that so hopefully this has been been

have been helpful in in sort of giving

you overview on why digital twins are

the most productive remote worker that

you can hire just make sure that you

hire it for for the right reasons and

how you can apply this technology to

have a massive impact and as we move

from the control room to the bedroom

thank you for your time really

appreciate this and we'll see you on one

of our future webinars

Last updated