馃攳
Azure Digital Twins demo | Creating replicas of real-world environments - YouTube
Channel: Microsoft Azure
[7]
- Hi, I'm Alyssa Sharp,
[9]
a senior program manager in Azure IoT.
[12]
I'm excited to show you how Azure Digital Twins
[15]
lets you create an intelligent environment,
[18]
combining data from all of your IoT devices,
[21]
business applications,
[22]
and really any other data source
[24]
to model, monitor and manage real-world environments.
[28]
For this demo,
[29]
we will show how Azure Digital Twins modeling capability
[33]
surfaces new insights
[35]
and keeps the supply chain of a clothing company on track.
[38]
As I go through this demo,
[39]
I challenge you to think and imagine how
[42]
the capabilities of Azure Digital Twins
[45]
can be easily transposed or applied to your industry,
[48]
your processes,
[50]
your environments,
[51]
and can ultimately keep your business operating smoothly.
[54]
Really, the possibilities are endless.
[56]
Let's dive in.
[59]
Here is the global view of your suppliers,
[61]
manufacturers, transportation,
[63]
warehouses and retailers.
[66]
By combining data from IoT sensors,
[68]
business applications,
[70]
third-party sources,
[71]
and applying the Azure Digital Twins model,
[74]
you get a real-time representation of the health
[77]
of your supply chain.
[78]
For example, if you want to look at the production in one
[81]
of your factories in South America,
[83]
you can click on the factory
[85]
and important metrics are given.
[87]
If you manufacture your own goods,
[89]
you're able to view what's happening within this factory.
[92]
The efficiency would track the energy consumption
[95]
and would pull data from your IoT sensors
[97]
so you can make sure that it stays within the level
[100]
the company has outlined.
[101]
The reliability pulls in data from your ERP system,
[105]
showing you if the factory ships on time,
[107]
with good quality and within budget.
[110]
Open orders will let you quickly see what
[112]
the factory is preparing.
[114]
An intelligent environment doesn't refer to
[116]
a standard building.
[117]
If I click into one of our shipments,
[119]
I can see the status of our goods and transports.
[122]
Tracking each container's location,
[124]
temperature and humidity.
[126]
I can quickly identify if a product will be damaged
[129]
before it even arrives to a retailer.
[132]
This particular shipment is scheduled to arrive on time
[135]
and is currently within the humidity levels
[137]
we need to maintain a good product.
[139]
Nobody wants any moldy clothes.
[141]
All of this data passes through Azure IoT Hub
[144]
and into Azure Digital Twins
[146]
where it's combined with all of the business data
[149]
you need to provide the cost and context.
[152]
Here, in the Azure Digital Twins Explorer,
[154]
you can see the model of the supply chain
[157]
created using our Digital Twins definition language.
[160]
Here we can see how the digital twins are connected
[164]
to ultimately create that intelligent environment.
[167]
Within each twin,
[169]
I have specific metadata captured and then surfaced
[172]
on the previous dashboard.
[174]
Here is the factory we looked at.
[176]
You can see the model on the right side of the screen
[178]
showing how the factory is performing.
[180]
We can see the reliability,
[182]
the efficiency and the open orders.
[184]
Same with the shipment.
[186]
Here is the digital twin of the shipment heading across
[188]
the Indian ocean with all of the associated metadata.
[192]
Going back to our global view,
[194]
we just received a message on Microsoft Teams
[197]
from one of our supplier relationship managers.
[199]
Their client is having a problem with a T-shirt shipment
[202]
in California.
[204]
Because the Azure Digital Twin model
[206]
is applied to every aspect of the supply chain,
[208]
we can click on the T-shirt product
[210]
and immediately narrow down the retailers for this shirt
[213]
in California.
[215]
Let's click on the retailer
[217]
and see their current inventory and past orders.
[220]
With Azure Digital Twins flexible modeling capability,
[223]
Each model you create will have unique attributes
[226]
and metadata.
[227]
We can see the retailers past order history
[230]
and find the one with the T-shirts.
[231]
With the historical data available
[233]
and the highlighted supply chain,
[235]
we can turn back the clock and see where
[237]
the issue might have occurred.
[238]
Let's check the transportation of the product.
[241]
By clicking this,
[242]
I can see the average humidity,
[244]
temperature and when the shipment was delivered.
[247]
So far, everything looks good.
[249]
Let's click into the factory and check
[251]
the temperature and humidity levels
[252]
during the time the product was created.
[255]
If they're off,
[256]
this could be a major issue for current clothes
[258]
being produced.
[259]
So let's see what's happening.
[261]
So we're moving from the global intelligence ecosystem
[264]
to an intelligent environment
[267]
and ultimately to the individual assets.
[270]
Within this factory,
[271]
we can monitor the production lines remotely
[274]
and in real time.
[275]
By clicking on each line,
[277]
I'll be able to see where the humidity is a problem.
[281]
And once we click on the third one,
[282]
we can see the temperature and humidity is high.
[286]
It looks like this might be the reason
[288]
the T-shirts were ruined.
[289]
Let's see how the line performed when
[291]
the T-shirts in question were created.
[293]
Using Azure Time Series Insights,
[296]
I can compare the performance of this line
[298]
versus the other lines at the same time
[301]
to see if the problem occurred previously.
[304]
Azure Time Series Insights and Azure Digital Twins
[307]
seamlessly connect.
[308]
The TSI Explorer shows me the historical performance
[311]
of all of the lines during the given time
[314]
this order was taking place.
[316]
It'll be most helpful to view these as a stacked graph
[320]
so I can quickly identify any anomalies.
[323]
Now, it's very apparent the humidity and temperature
[326]
of this line has been an issue for a while
[329]
and we need to take immediate action.
[331]
With the Azure Digital Twin model in place,
[334]
we can see the product lines and orders
[336]
impacted by this particular factory.
[339]
The California retail store has about $20,000
[342]
in impacted inventory.
[344]
The UK store has about 12,000.
[347]
We can't afford to have any more shipments ruined,
[349]
so we're going to turn this factory offline
[351]
until it's able to be fixed.
[353]
Using machine learning,
[355]
the system can re-route orders from factories
[357]
that have the capacity
[359]
and the product lines are able to turn green again.
[362]
Azure Digital Twins makes it possible
[364]
to connect data from sources throughout your entire business
[367]
and apply a model
[368]
so the data is contextualized
[370]
and you can get insights faster.
[373]
The flexible modeling capabilities allow you
[376]
to create custom models that apply to your individual
[379]
and business needs.
[381]
And because it's built on Azure,
[383]
we offer the scalability and security of
[385]
the entire Azure platform.
[387]
If you'd like to try this demo out for yourself,
[389]
as well as play around with the Azure Digital Twins models
[392]
and explore,
[394]
visit our GitHub repository at aka.ms/IoTDemos/supplychain.
[402]
Thanks for watching.
Most Recent Videos:
You can go back to the homepage right here: Homepage